5
Energy Management in LTE Femtocells

Kapil Kanwal1, Ghazanfar Ali Safdar1, Masood Ur Rehman1 and Xiaodong Yang2

1 School of Computer Science and Technology, University of Bedfordshire, Luton, UK

2 School of Electronic Engineering, Xidian University, Xi’an, China

Wireless cellular networks have seen dramatic growth in number of mobile users. As a result, data requirements and hence base‐station (BS) power consumption has increased significantly. In turn, this adds to operational expenditures (OPEX) and also causes global warming. The BS power consumption in Long Term Evolution (LTE) means, therefore, it has become a major challenge for vendors to stay green and profitable in a competitive cellular industry. It necessitates novel methods to devise energy efficient communication in LTE. Importance of the topic has attracted huge research interests worldwide. Energy saving (ES) approaches proposed in the literature can be broadly classified in categories of energy efficient resource allocation, load balancing, carrier aggregation and bandwidth expansion. Each of these methods has its own pros and cons leading to a trade‐off between ES and other performance metrics resulting in open research questions. This chapter discusses various ES techniques for the LTE systems and critically analyses their usability through a comprehensive comparative study.

5.1 Introduction

Wireless communication has become one of the basic provisions of the modern world. Since the inception of first radio communication system by Marconi [1], wireless communication systems have seen a massive growth in the last few decades from having a couple of individuals to the majority of the world as their users [2, 3]. The concept of frequency reuse was first introduced in cellular radio communication systems by AT&T [4]. Further developments in radio communication introduced digital cellular systems, which pass through a long chain of evolution known as generations (G). We have seen usage of 1G, 2G, 3G and now 4G as the communication standard with each resulting in enhanced performance of cellular systems [5, 6]. Aiming for key achievements such as short transmission time, high throughput, low latency and security [7, 8], these systems generally consist of BS connected to core network. Each BS has designated coverage area, called cell and communicates directly with User Equipment (UE) within its coverage [9–11]. Whenever UE moves from serving cell to neighbour cell, its transfer of control is initiated through handover process [12, 13]. LTE is a 4G technology that transmits Digital Broadband Packets over Internet Protocol (IP) while offering a peak data rate of 100–300 Mbps [14–16]. This increased data rate in LTE is achieved by employing Orthogonal Frequency Division Multiple Access (OFDMA) based technology that promises low latency, high data rate and packet optimized radio access [17]. This enhanced performance of services compared to previous generations of the cellular networks has helped LTE systems to gain rapid popularity both commercially and academically.

5.2 Architecture of LTE Networks

Since the chapter is focused on energy management in LTE networks, this section presents a brief overview of the LTE architecture. LTE systems usually provide low latency, high data rate and packet optimized radio access. Compared to 3G, LTE additionally provides international roaming and compatibility with other legacy networks [18, 19]. The 4G systems make use of OFDMA and Single Channel Frequency Division Multiple Access (SC‐FDMA) schemes to support flexible bandwidth [20–26]. LTE architecture is generally based on Evolved Packet Core (EPC), Universal Terrestrial Radio Access (UTRA), and Universal Terrestrial Radio Access Network (UTRAN), each of which communicates with core network air interfaces and radio access network [27, 28].

Figure 5.1 illustrates the overall architecture of LTE networks showing both EPC and evolved UTRAN (E‐UTRAN) [29–31] while Table 5.1 summarizes the core elements of LTE architecture.

An oval divided into 2 by a horizontal line, with upper part labeled EPC and lower part E-UTRAN. The upper part has 3 linked boxes labeled S.GW, HnodeB GW, and MME and the lower part has 4 linked towers labeled eNodeB.

Figure 5.1 Architecture of LTE networks.

Table 5.1 LTE network components.

Components Description
Evolved Packet System (EPS) Provides IP connectivity using E‐UTRAN.
Mobility Management Entity (MME) Responsible for authorization, security, handover, roaming and mobility of users.
S1 Interface It connects EPC with BSs.
Serving Gateway (S.GW) EPC terminates at this node. It is connected to E‐ UTRAN through S1 interfaces. Each user is allocated unique S.GW which is responsible for handover, packet routing and forwarding functions.
Packet data network gateway (PDN‐GW) PDN‐GW provides UEs with access to packet data network by allocating IP addresses. It is also responsible for secure connection with untrusted devices from non‐4G networks.
HnodeB Femtocells that are employed to improve seamless connectivity in coverage holes.
eNodeB Also known as BS that serves the UEs.
HnodeB GW Provides connection to the core network.
X2 Interface Provides communication between two BSs.

5.2.1 Communications Perspective Challenges in LTE Networks

Though LTE has proven to be a promising technology, it is a complex network and there are some challenges that need to be carefully addressed for optimum functionality.

5.2.1.1 Signalling System

In LTE networks, one of the major issues is to avoid or limit signalling overhead and overlapping in control part of the network. A large number of connections between nodes and network fragmentation causes a rapid increase in signalling traffic. Any failure in signalling system will drag operators towards increased system latency and outages resulting in loss of revenue [32, 33]. Increased signalling traffic also leads to increased energy consumption and definitely needs to be looked in carefully.

5.2.1.2 Backward Compatibility

LTE is usually compatible with all other relevant major standards. The combination of devices, network interfaces and equipment to support other standards complicates end‐to‐end functionality testing and interoperability testing (IOT) [34, 35].

5.2.1.3 BS Efficiency

Due to the employment of OFDMA in LTE, signals have high amplitude variability known as Peak‐to‐Average Power Ratio (PAPR), which reduces transmitter efficiency. Furthermore, the BS provides high data rate at the cost of high dynamic transmission power. Since, high transmission power results in increased energy consumption and thereby increases OPEX; energy management has become major challenge in LTE networks to stay profitable and also to reduce global warming [36].

5.2.2 Importance of Energy Management in LTE Networks

Since increased power consumption (using non‐renewable energy sources) directly contributes in climate change, therefore it has become major obstacle for environmental and economic aspects [37, 38]. Vendors highlighted the raising trend of power consumption due to the increased data traffic. Number of users of the mobile networks has 10% annual increase across the world with an increase of 25–50% in each user’s data requirements [39]. Therefore, provision of high data rate demanding services with minimum power consumption has become a major challenge for vendors to stay profitable [40, 41]. Information and Communication Technology (ICT) contributes approximately 10% of worldwide power consumption while it adds 2% to global warming [42, 43]. Moreover, global warming is growing swiftly due to the additional advance BSs being deployed to fulfil increased users demand thus resulting in a 15–20% increase per year and this increase almost becomes double every 5 years [44]. Noteworthy, ICT contribution in global warming will become 3% by 2020 [45]. Since, BS consumes major part of energy in LTE networks, reducing power consumption at BS could help cutting down OPEX [46]. Vendors choose to deploy automated networks to facilitate ES [47]. 3GPP has already introduced Self‐Organized Networks (SON), which increase the level of automation achieved in operation and maintenance, thereby resulting in a decreased OPEX [48]. Apart from other functionalities, SON also provides opportunities for incorporation of enhanced ES techniques that can help achieve reduced OPEX values. Technologies based on the concept of SON (e.g. LTE), can enjoy a 19% reduced OPEX due to advanced ES techniques [49].

5.3 Classification of ES Schemes

The literature presents considerable amount of research work on energy efficiency in LTE systems. Each BS in cellular networks consists of Baseband Units (BBU) with one or more transceivers. Each transceiver contains Radio Frequency (RF) part, Power Amplifier (PA) and Antennas connected through cables [50–52]. All these components are located very close to each other in a unit called Radio Resource Unit (RRU). PA is the main power‐hungry element in this unit [53]. Since LTE employs OFDMA [54–56] and normally PA operates at a level that is 6–12 dB lower than the saturation point, this results in lower adjacent channel interference. Power consumption at BS can be categorized as static and dynamic power consumption [57]. Static power consumption belongs to hardware used in BS and remains nearly constant. Dynamic power (also known as communicational power) on the other hand, depends on traffic load between BS and UEs [58, 59]. The focus of this chapter is to investigate, classify and critically analyse existing ES techniques to control the dynamic power consumption.

5.3.1 Static Power Consumption

Static power is purely hardware based constant power consumption, which the BS needs to cater for necessary operations. The static power consumption can be improved by energy efficient hardware designs and subsequent intelligent deployments. However, in this chapter our main focus based on dynamic power consumption.

5.3.2 Dynamic Power Consumption

The dynamic power consumption depends on the BS’s resources utilization and is directly affected by BS transmission operations. Therefore, it could be reduced by turning off of BS operations during idle states. Dynamic power management has attracted attention of researchers and could be classified in to three main categories, that is, energy efficient resources allocation, bandwidth expansion and load balancing as shown in Figure 5.2.

Oval labeled BS power consumption branching to static power and dynamic power consumptions. Dynamic power consumption branches to energy efficient resource allocation, bandwidth expansion, and load balancing.

Figure 5.2 Classification of energy saving schemes.

The dynamic (or communicational) power consumption can be reduced by appropriate activation and deactivation of BS’s transceivers also known as Discontinuous Transmission (DTX) during off peak time periods. DTX based schemes allocate Multicast and Broadcast Single Frequency Network (MBSFN) subframes through traffic load consideration [60, 61]. The power aware algorithm analyses the traffic that cells need to serve, then calculates the amount of resources required and distributes them among the subframes (SFs) to minimize the power consumption. In lightly loaded conditions, there is a possibility of some frames being not utilized, which could help to achieve improved energy conservation by configuration and turning off of idle frames. Importantly, micro cell DTX significantly reduces power consumption during low traffic rate; however, it does not work during high traffic loads because no empty SFs are left. Results indicate that average power consumption per cell without DTX is 170 W [60], while micro DTX enabled cell reduces this figure by 60%. Next to this, the work presented in [62–64] avail the concept of energy efficient resource allocation and significantly reduce overall BS power consumption. Aggregation of resource blocks through carrier aggregation algorithms can also help achieve better ES [65, 66]. This helps in reducing overheads, thereby increasing energy conservation. Along the same line, energy efficient carrier aggregation algorithms group together the component carriers (CC) to achieve greater ES [67]. Distance aware schemes, which involve switching off the BS that is greater distance from UEs, can also help to obtain better ES [68–70]. These schemes reduce energy consumption by appropriate activation/deactivation of the BS, based on information of varying distance and load. Another dynamic traffic‐aware approach is introduced in [71], which uses time varying traffic information for energy conservation. Each BS divides its cell in different number of sectors, then switch off appropriate sector (with low traffic) providing power saving opportunities. Centralized and distributed schemes that engage UE migration also help achieve better ES [72, 73]. Centralized schemes select the highest loaded BS through analysing traffic information and determine if it could accommodate more UEs. Considering selected BSs, if available bandwidth is greater than the capacity required to serve neighbour cells, UEs with lowest load traffic will then be shifted towards heavily loaded BSs, resulting in switching off lightly loaded BSs for reduced energy consumption. Compared to centralized, distributed schemes in contrast, it is possible to select a pair of BSs and then determine the ES state of each BS. Initially, schemes activate ES on particular BS, which examines the neighbour cells list, and select one BS with the lowest load forming a pair. The BS preferring to keep powered ON is the one that can accommodate more UEs. On the same lines, work presented in [74–76] employs distributed scheduling and energy efficient power control approaches for reduced power consumption. Another study in [77] shares a relay between different operators, thus resulting in ES. In [78] authors introduced an energy efficient link adaptation scheme that combines the traditional link adaptation with power control, thereby resulting in improved energy efficiency at the BS. This scheme uses the BS’s transmitted power as a new feedback parameter and predicts an optimal set of parameters in order to maximize the BS’s energy efficiency and satisfy the Block Error Rate (BLER) constraint for the channel state. Another interesting scheme is presented in [79, 80], which suggests an energy efficient resource allocation scheme that operates in multi‐cell OFDMA‐based LTE networks. In the same context, [81, 82] introduce energy efficient resource allocation for reduced power consumption in LTE networks. This method combines dynamic Resource Block (RB) allocation with energy efficient power allocation and reduces overall BS power consumption. A bandwidth expansion scheme with load balancing is introduced in [83, 84], which employs the idea of moving UEs from overlapping area of lightly loaded cell to the heavily loaded cell. The Time Compression Mode (TCoM) is presented in [85], which saves power by reducing control signals overhead’s transmission. RBs are compressed together in TCoM, either in time or frequency domain by usage of higher order modulation. ES is achieved through reduction in overhead signalling by appropriately turning off of the unused RBs [85]. On the same lines, optimized resource allocation could also lead to reduced energy consumption as discussed in [86]. Energy efficient BS deployment too has helped in improved energy conservation [87]. ES approaches for D2D communication in LTE networks resulting in reduced power consumption are presented in [88–90]. The most recent research work has established the idea of integrating Cognitive Radio Networks (CRNs) [91] with LTE infrastructure for improved ES. This predominantly lies in the fact of isolating users in two categories (i.e. PUs and SUs). During awake periods, BS transmits PUs data over the licensed spectrum while in contrast, SU data is sent over the unlicensed spectrum. BSs are switched in to sleep mode right after completion of data packets transmission thus resulting in opportunities for improved power conservation. ES can also be achieved through suitable cells coverage expansion and turning off of idle BSs [92]. Importantly, this scheme initially splits cells in two main categories; that is, cooperative cells and dormant cells. Where, cooperative cells serve associated users while dormant cells are turned off during low traffic time periods for ES. Next to this, intelligent resource allocation and power control [93, 94] can help reduce dynamic power consumption thus resulting in improved energy efficiency. It is worth noting that energy efficient schemes, while deployed at every BS, allocate lower transmit power to suitable resources in line with the associated Signal‐to‐Noise plus Interference (SINR) ratio. Among others, the D2D communication based scheme presented in [95] uses energy efficient heterogeneous routing for enhanced energy conservation. A significant amount of research work has been carried out to develop different ES schemes, which help to reduce dynamic power consumption. However, the increasing trends of OPEX and global warming indicate that there is always a need to do more research work to achieve enhanced ES for future wireless systems. Based on the previous discussion, a broad classification of ES schemes is presented in Figure 5.2 while detailed insights into individual schemes are provided in the following sections.

5.4 Energy Efficient Resource Allocation

In order to transport UE data across wireless media, wireless cellular systems employee various control channels that segregate dissimilar types of data and transport them across Radio Access Network (RAN) in an orderly routine. LTEs consist of physical channels, transport channels and logical channels. Further, physical channels consist of Physical Broadcast Channel (PBCH), Master Information Block (MIB) and Physical Downlink Control Channel (PDCCH). Intelligent switching on and off of these control channels can result in increased ES. Some energy efficient schemes in energy efficient resource allocation category are explained next.

5.4.1 Hybrid FBS and MBS Based Schemes

The use of femto base stations (FBS) has proven to be promising technology to cover those areas where macro base stations (MBS) are limited. In the same context, work presented in [93] introduces power control based RBs allocation scheme in LTE network with MBS and FBS, which employ the concept of Almost Blank Subframe (ABS) and Reduced Power Blocks (RPBs) to allocate reduced transmission power to RBs, thereby resulting in to reduced downlink power consumption. This scheme is recommended for two tier heterogeneous networks with MBS and few FBSs as shown in Figure 5.3. The main idea lies in the fact that varying transmit power levels can be assigned to different resources thus resulting in reduced BS power consumption. The level of transmit power is measured through SINR, thus if users SINR is higher than the predefined threshold, then they are allocated with lower transmit power, while higher power is assign to users with lower SINR. Since SINR values changes rapidly, accordingly estimation of transmit power also changes continuously. Next to this ‘Breathing’, a technique is introduced for RBs allocation that divides users in two classes; that is, Inhale and Exhale. Users are arranged in ascending order in the Inhale class in relation to the required transmit power and are mapped with RBs in sequence. On the other hand, the Exhale class involves sorting users in descending order of their transmission power value [93].

An oval with 2 FBS, 1 MBS, and 6 users (A–F). Arrows from MBS and FBS point to a box with transmit power scale for users A–F. The box is connected to a circle with texts, “Users are arranged in ascending/descending….”

Figure 5.3 Hybrid FBS and MBS based ES scheme.

5.4.2 Link Adaptation Schemes

LTE provides a high data rate through the effective resources utilization in available bandwidth. The Channel Quality Indicator (CQI), Precoding Matrix Indicator (PMI) and Rank Indicator (RI) parameters play key roles in efficient use of resources. PMI determines which precoding matrix should be used for downlink transmission while RI presents the number of layers that should be used for downlink transmission. CQI is reported from UEs to the BS that contains information about the supported Modulation and Coding Schemes (MCS). CQI plays a major role in selection of MCS at downlink in BS. CQI values range from 0 to 15. A higher value of CQI indicates use of higher modulation scheme while BS can use higher coding rate for increased energy efficiency [78]. An energy efficient (EE) link adaptation scheme, which combines traditional link adaptation with power control resulting in improved energy efficiency at BS is presented in [78]. This scheme uses BS transmitted power as a new feedback parameter and predicts optimal parameters that maximize the BS energy efficiency and satisfy the BLER constraint used for demodulation tests in multipath conditions during radio link monitoring. This scheme can be best described with the help of the LTE based downlink transmission model shown in Figure 5.4. UE estimates channel gain between BS and itself to calculate the parameters RI, PMI, CQI and transmit power. These parameters are then fed back to BS through the feedback channel as shown in Figure 5.4. The BS uses feedback received from UEs as an input parameter to adjust its transmission power; where RI helps to determine the code word, CQI helps to select MCS scheme for each transmission and PMI is used by the BS to select the precoding matrix.

Diagram of link adaptation scheme of LTE based downlink transmission with boxes labeled Turbo coding and QAM modulation, Layer mapping, Precoding, Resource element Mapping, OFDM signal generation, etc.

Figure 5.4 Link adaptation scheme – LTE based downlink transmission.

5.4.3 Cross Layer Resource Allocation Schemes

A cross layer based EE resource allocation scheme for multi‐cells OFDMA‐based LTE networks is presented in [79]. This technique encompasses physical and medium access control (MAC) layers combining dynamic RB allocation at MAC layer with EE power allocation at Physical Layer (PHY), thus resulting in reduced overall power consumption by the BS. Dynamic RB allocation is based on feedback (energy efficiency indicator) that is used to adjust scheduling process. This method also promotes the user’s fairness through allocating equal resources to all users either with good and bad quality channels.

5.4.4 MBSFN Resource Allocation Scheme

The MBFN predicts the future traffic load that needs to be served in the next frame, this predicted load is used to calculate the required RBs while turning off the unused resources. The future load prediction is made using previously served load information exchanged between BSs through X2 interface (standard interface used for BS communication in LTE). An interesting MBSFN based ES scheme in [61] configures MBSFN SFs that help to provide and setup transmitter switching off periods. Additionally, this method estimates the resources required to serve the predictive load for effective resource allocation resulting in enhanced power saving by turning off the idle resources. Based on LTE specifications, 6 out of 10 SFs can be configured as MBSFN (Figure 5.5). Importantly, MBSFN SFs carry fewer Reference Signals (RS) compared to the standard subframe. Therefore, in a case where no data is available, MBSFN SFs can be turned off resulting in reduced energy consumption.

MBSFN based frame architecture with 2 sets of MBSFN sub-frames and 4 active sub-frames. On the right is a 12 by 6 grid with a callout with text, “Each resource block contains data part and control signals part.”

Figure 5.5 MBSFN based frame architecture.

5.5 Bandwidth Expansion Schemes

The EE LTE networks can also be realized through bandwidth expansion. Several proposed techniques employing bandwidth expansion for improved energy efficiency are presented here.

5.5.1 CoMP Based Coverage Expansion

Work in [92] uses Coordinated Multiple Point (CoMP) for improved ES. CoMP expands cell coverage thus resulting in better expansion compared to antenna adjustments and transmission power measurements (see Section 5.5.6). The proposed work employs link budget and SINR as input parameters and then divides networks in clusters on the basis of the equivalent cell principle with a distributed method (Figure 5.6). Cells in this scheme are divided into two main categories, that is cooperative and dormant cells, which is decided by a Joint Processing (JP) cooperative cell selection model. During off peak traffic time periods, cooperative cells expand their coverage to serve dormant cells that are turned off for ES purposes.

A box labeled network with differently shaded hexagons marked by circles and oval labeled clusters. A right arrow from the box points to 2 dashed hexagons (dormant cells) and a solid hexagon (cooperative cell).

Figure 5.6 CoMP based coverage expansion.

5.5.2 Time Compression (TCoM) Scheme

The 10 ms frame in OFDMA consists of 10 SFs. Each SF includes two slots of 0.5 ms each and each time slot consists of 12 subcarriers and seven symbols as shown in Figure 5.7. Subcarriers of each symbol can be allocated to multiple users thereby making efficient use of radio resources. TCoM tries to reduce the power consumption caused by the usage of higher order modulation schemes in OFDMA through a decrease in control channel overhead [85]. RBs in TCoM are compressed together and ES is achieved through reduced overhead signalling by appropriately turning off unused RBs during the idle state. The time and frequency implementations of TCoM do not differ in performance because of the fact that changes in either length or bandwidth of a transmission have the same impact on the transmitter’s energy. A compression factor to represent the number of RBs to be pooled together is introduced in [85]. It also uses Shannon’s capacity to derive required Signal‐to‐Noise Ratio (SINR). TCoM is found to be around 26% more EE compared to the LTE benchmark standard.

Illustration of OFDMA frame architecture, with 12 by 10 grid with shaded area labeled one resource block. Above the grid are horizontal bars representing 1 frame, 10 subframes, 2 slots, and 1 resource block.

Figure 5.7 OFDMA frame architecture.

5.5.3 Bandwidth Expansion Mode (BEM) Scheme

Another Bandwidth Expansion Scheme (BEM) is described in [83]. This method is based on the concept that when the network is lightly loaded (larger number of RBs is free), in this scenario bandwidth allocation can be increased to reduce power consumption at BS. In LTE systems, minimum resource allocation is one RB for each user and allocation is done by schedulers. Expanded RB allocation (allocating more than 1 RB per user) reduces the MCS and SINR per frequency channel for each user, which in turn provides more EE transmission. Work in [83] is especially recommended for low loaded networks, because extra RBs that are idle during off peak traffic helps in bandwidth expansion. BEM addresses two important factors; EE and Mobility Load Balancing (MLB) in networks. This work proposes an effective EE resource allocation optimization model by employing a low complexity method called Energy Efficient Virtual Bandwidth Expansion Mode (EE‐VBEM). The concept of Virtual Load Balancing (VLB) that distributes some of the traffic (users) from highly loaded cells to the lightly loaded cells is used as shown in Figure 5.8. The EE‐VBEM consists of two major parts: (1) EE Resource Allocation Optimization Model and (2) Low Complexity Method to achieve (1). Firstly, all BS exchange load information of neighbouring cells through the X2 interface. Based on this information, each BS determines whether there is a need for load balancing. In case load balancing is required, VLB automatically start shifting users from overlapping area to lightly loaded BS. BEM then calculates the required RBs for each UE using minimum required data rate and user channel quality. Once RB calculation is done, the BEM prioritizes the users according to SINR value. A higher SINR indicates higher BEM priority for the user and vice versa. After priority assignment, RBs are allocated to the UEs. BEM saves energy by allocating extra resources at the expense of reduced overall capacity of the BSs [83].

2 Overlapping circles with towers labeled eNodeB A (left circle) and eNodeB B (right circle) for Node A and Node B users, respectively. The intersection is labeled User of node B is shifted to node A.

Figure 5.8 Resource allocation through load balancing.

5.5.4 Component Carrier Based Schemes

Carrier aggregation is a well‐known technology used in LTE networks for the effective use of bandwidth. Each aggregated carrier is known as a Component Carrier (CC) that can have bandwidths ranging from 1.4, 3, 5, 10, 15 or 20 MHz, while a maximum of five carriers can be aggregated at a time. Carrier aggregation can be achieved through three methods as shown in Figure 5.9. The simplest method is known as Intra‐Band Contiguous, which uses contiguous carrier aggregation at the same frequency band. The second method is known as Non‐Contiguous Intra‐Band Carrier Aggregation in which the CC operates at the same frequency band but has gaps as shown in Figure 5.9. The third method is Non‐Contiguous Inter‐Band Carrier Aggregation in which carriers operate at different frequency bands. To achieve EE communication in LTE networks, more CCs can be jointly utilized in a BS for enhanced ES opportunities. In [65], authors recommended OFDMA‐based multiple CC technique for EE transmission that uses two CCs for data transmission. The main idea is to transmit only necessary CCs thus providing opportunities for appropriate deactivation of idle CCs to reduce the power consumption. The ES scheme in [65] works in the downlink in BS and supports both real and non‐real‐time traffic simultaneously as shown in Figure 5.10. The ES scheme consists of two CCs operating at same frequency band and can be jointly utilized in BS for data transmission. The two CCs are called the Primary Component Carrier (PCC) and Secondary Component Carrier (SCC), respectively. Normal data transmission uses PCC while SCC is only used during high traffic conditions. During transmission, a user’s data packets are transmitted to the session level where they are classified as Real Time (RT) or Non‐Real‐Time (NRT) by the classifier and forwarded to RT and NRT Queues, respectively (Figure 5.10). The data packets then wait in transmission queue to be served by the proposed ES scheme, which consists of two algorithms. The first algorithm allocates radio resources, while the second algorithm is used for the appropriate activation/deactivation of the SCC. The first algorithm further contains two sub‐algorithms; Bandwidth Allocation Algorithm (BAA) and Resource Block Allocation Algorithm (RBAA), respectively. All these algorithms are executed at the beginning of every SF and jointly provide ES opportunities at the BS.

3 sets of 4 connected boxes for contiguous intra band carrier aggregation of 4 × 20 MHz component carriers (top) and non contiguous intra band carrier aggregation of 3 × 20 MHz component carriers (middle, bottom).

Figure 5.9 Carrier aggregation.

Illustration of OFDMA‐based CC ES scheme, with arrow from a box labeled Session start pointing to Classifier, to real time and non real time queue, to transmission queue, then to PCC and SCC 2 RBs in subframe.

Figure 5.10 OFDMA‐based CC ES scheme.

5.5.5 Scheduling Based Schemes

Videv et al. have presented an EE scheduling scheme in [96]. The method is based on bandwidth expansion through allocation of extra resources to the UEs and uses lower order modulation schemes for ES. This scheme reduces power consumption by 44% while maintaining throughput and QoS constraints. It uses an energy‐aware scoring scheduler, which considers best channel conditions and allocates additional resources to the UEs. The scheduler allocates resources by following the integer factor defined for bandwidth expansion. This method is effective only for networks where traffic is low and more free resources are available to be allocated to the UEs. This scheme provides ES at the cost of overall system capacity and therefore is not effective in a practical RT environment.

5.6 Load Balancing Schemes

Research has shown that traffic load varies significantly at the BSs and a lot of energy is wasted during low load operation. Load balancing is a part of Radio Resource Management (RRM). The term load balancing presents any method that could be used to transfer load from highly loaded cells to lightly loaded neighbour cells for the efficient use of radio resources. The user’s distribution and traffic flow are irregular in cells, which can cause an unbalanced load condition in the network. In wireless cellular networks with unequal traffic load distribution, some of the users at the edges of cells can be transferred from highly loaded cells to the lightly loaded cells thereby providing opportunities for efficient resources utilization. When UEs detect that neighbour cells can provide better signal quality than its current serving BS, they are handed over to that neighbour cell. During load balancing, if the cell is desirable or already in ES mode and it is selected as a candidate for load balancing from a nearby heavily loaded cell, then two options exist. Firstly, to prioritize the load balance without considering the ES and, secondly, focus is made only to prioritize the ES. In the second case, UEs are not allowed to be handed over to the cells, which are desirable, or already in ES mode and the heavy loaded cell has to find another neighbour cell for load balancing. In this case, edge users may not be served efficiently but power saving could be improved.

5.6.1 Distance Aware Schemes

Work in [68] has introduced distance aware schemes that involve switching off a BS that is a greater distance from UEs. This work reduces energy consumption by appropriate activation/deactivation of BS through information of varying distance and load. Each BS in a seven‐cell‐based cluster calculates its average distance from associated UEs and adjacent cells UEs as shown in Figure 5.11. Since the larger average distance between BS and UEs leads to the higher power consumption, the appropriate BS (with greater average distance) is selected for switching off. If the bandwidth requirements to serve associated UEs are less than the total available capacity supported by adjacent cells, then the selected BS is switched off and traffic is allocated to the neighbour cells resulting in reduced power consumption. Moreover, the BSs in sleep mode can be activated if the network becomes busy due to high volumes of traffic. The ES scheme aims to turn off as much BSs as possible without any degradation of QoS. This scheme divides the day into two zones, a night zone (7 pm to 7 am) and a day zone (8 am to 6 pm). Turning off the BS is performed in the night zone to achieve ES during 12 hours of low traffic load conditions. The BS is switched on in the day zone when traffic load increases and network becomes busy. In high traffic load conditions, a number of BSs should remain switched on in order to serve the UEs appropriately without affecting the QoS. The ES scheme proposed in [68] significantly reduces power consumption by deactivating BSs while neighbour cells can send activation instruction back to the BS in sleep mode through an X2 interface.

A cluster of 7 hexagons with signal towers and cell phones (users). The towers are connected by lightning bolt symbols (X2).

Figure 5.11 Distance aware based BS communication.

5.6.2 Coverage Expansion Based Schemes

A centralized ES algorithm is proposed in [72] that provides ES by turning off the lightly loaded BSs. This scheme is based on the idea of shifting the traffic towards the highest loaded BS using load and coverage information of the network and switches off a lightly loaded BS. The main idea lies in the fact that all UEs of a lightly loaded BS are served by the neighbouring busiest BS, thus it permits a lightly loaded BS to switch off for ES. Initially, a neighbour BS sectorizes its coverage, then extends coverage of the appropriate sector through transmission power adjustment and reconfiguration of the antenna as shown in Figure 5.12. The extended sector coverage helps BSs to serve UEs where a lightly loaded BS is turned off. The proposed algorithm, while deployed at every BS, sectorizes and extends its coverage for ES purposes. It uses two algorithms; first, one monitors network for load information while a second operates on individual BS and manages its sectorization and transmission expansion process. Initially, on the basis of load information, a centralized algorithm selects the busiest BS and analyses its resources availability. If the selected BS has enough resources to serve the neighbour BS’s users, then one of its sector transmission coverage is expanded to serve UEs of the neighbour cell that is being switched off for ES as shown in Figure 5.12 [72].

Schematic displaying a neighbour cell coverage (top left) and highly loaded cell coverage (top right) with arrows pointing to a centralize ES based coverage expansion during sleep mode with 2 boxes containing texts.

Figure 5.12 BS coverage expansion for ES.

5.6.3 Distributed Schemes

In [71], a distributed self‐organized sectorization of BSs is presented for EE communication. Based on the varying load information, each BS reconfigures itself in RT, thus utilizing the minimum number of sectors for ES while promising adequate QoS. Since each BS dynamically reconfigures itself and no correspondence is required with neighbour BSs, this scheme is inherently distributed and self‐organized. Each BS is implemented with traffic‐aware algorithm for continuous reconfiguration of sectors depending on time varying load. Objective of traffic‐aware algorithm is to regulate sectorization and minimize the number of sectors in each BS while maintaining the necessary signal power required for each UE and QoS constraints. It is worth noting that during low traffic durations, a lower number of sectors is sufficient to serve BS users, thus unused sectors are turned off to achieve ES as shown in Figure 5.13. This scheme estimates the required number of RBs for each UE and uses both time‐inhomogeneous and time‐homogeneous traffic models for performance analysis. It also employs interference‐managing arrangements to handle inter‐cell interference and significantly reduces overall system dynamic power consumption by turning off the unused sectors in each BS (Figure 5.13). However, one of the major disadvantages of this technique is that a sector can be turned off only if it does not serve even a single UE in low traffic durations.

Illustration of distributed schemes displaying a hexagon divided into 6 equal parts for sectors 1–6 (left) and another hexagon divided into 3 parts for sectors 1–3 (right). Both have a tower in the middle.

Figure 5.13 Distributed schemes: sectorization in BS.

5.6.4 Shared Relay Based Schemes

Researchers have proposed a shared relay based load balancing ES scheme for the LTE networks in [77]. The operators or service providers share their network resources to accommodate additional users and support their demand of increased voice and data services through load balancing. This scheme, however, needs reasonable investment in the network infrastructure and is based on two assumptions. First assumption states that two different network operators jointly provide coverage to the service area through service level agreement, which allows UEs to communicate with operators through a load balancing algorithm. A second assumption says that a centralized SON algorithm is used for optimization of communication between UEs for ES. It lays the foundation for a relay based shared network based on two LTE networks belonging to two different operators with their own BSs. UEs from both operators can freely communicate with any BS regardless of their operator. BSs of both operators are connected through backhaul link, which is monitored and controlled by a remote entity called a RAN. Using information about the load and channel conditions, the SON algorithm calculates the Reference Signal Received Power (RSRP) of both BSs for each user. Once the RSRP has been calculated, the UE is then allocated to the BS having better RSRP for it. However, if the RSRP of both BSs is the same then UEs prefer to communicate with their own operator BS because both operators prefer to utilize their own resources first. This scheme reduces power consumption by 15–20% with the help of SON based load balancing.

5.6.5 CRN Adopted Switching Off of a BS

The work presented in [91] incorporates CRNs with LTE and turns off BSs for ES purposes. The proposed algorithm employs three modes of operation, namely sleep, awake and listening modes, respectively (Figure 5.14). During the awake mode, PU data is transmitted using pre‐emptive priority while SU data is sent using unused remaining spectrum. Once all packets have been transmitted and the buffer becomes empty, the BS is turned in to sleep mode for ES purposes. Importantly, arrival of PU data can shift a BS from sleep mode straight back to awake mode. Otherwise, the BS remains in sleep mode to conserve energy and shifts to the listening mode upon expiry of sleep mode timer. The BS listens to data traffic in listening mode before it repeats the whole cycle.

State diagram for CRN based ES displaying a circle with 3 ovals for sleep mode, listening mode, and awake mode that are linked by arrows. Boxes with text are found alongside of the arrows.

Figure 5.14 State diagram for CRN based ES.

5.6.6 Reduced Early Handover (REHO) Scheme

Taking into account challenges and open research issues, we have proposed a reduced early handover (REHO) energy scheme in [97]. REHO merges bandwidth expansion with RBs switching off for enhanced ES purposes. The REHO scheme, while deployed at every BS, relocates users from overlapping areas of seven neighbour cells to the one centre cell through load balancing thus enabling neighbour cells to turn off freed RBs for ES. REHO employs the concept of time compression, thus combining two RBs to form one incremental RB and allocates to a single user resulting in reduced control channel overhead transmission; further, it initiates early handover using a reduced value of hysteresis. REHO achieves ES through fine‐tuning of hysteresis, offset and is explained with help of Figure 5.15. A BS transmits cell specific Reference Signals to all users within its coverage area, which are used by users to measure RSRP. When the RSRP of a target cell becomes better than the serving cell, then the user triggers an A3 event and sends a measurement report (best candidate BS information) to the serving cell to initiate handover [98]. The hysteresis and offset are used to push a user closer to the target cell thus ensuring minimal radio link failure. REHO uses the minimum reduced value of hysteresis thus resulting in an early handover initiation compared to standard handover for ES while maintaining radio link failure at acceptable levels. Systems level simulations are performed to demonstrate the behaviour of REHO. The chosen network scenario consists of seven overlapping cells with 50 users randomly distributed in each cell. Figure 5.16 compares REHO with standard LTE handover for dynamic power consumption. Clearly, REHO outperforms standard handover in terms of dynamic power consumption where the reduction in power consumption in REHO is achieved by early turning off of RBs.

REHO ES scheme with 2 overlapping ovals labeled serving BS and target BS, each having a tower in the middle, with a car in the intersecting area. Various lines represent BS reference signal, user mobility direction, etc.

Figure 5.15 REHO ES scheme.

Plot of dynamic power consumption vs. time having 2 curves with solid dots and star markers representing proposed scheme and benchmark, respectively.

Figure 5.16 REHO dynamic power consumption.

5.7 Comparative Analysis

Table 5.2 critically compares existing ES schemes in terms of their pros and cons followed by detailed discussion and analysis. Table 5.2 shows that distance aware scheme [68] operates during 12 hours and save energy up to 30% as compared to always on network during night zone. Since traffic load is high in the daytime, distance aware schemes fail to turn off BSs during day time and are only effective in the night zone when traffic load is low. The dynamic distance aware approach achieves ES up to 70% in comparison to an always‐on network and operates every hour in contrast to the distance aware scheme [69].

Table 5.2 Critical analysis of existing ES schemes.

ES Schemes Advantages Disadvantages
Distance Aware [68]
  • Power saving up to 30% compared to an always ON network.
  • Runs during a limited time period (8:00 pm – 8:00 am).
  • No power saving during the peak traffic time period.
Dynamic Distance aware [69]
  • Power saving up to 70% compared to an always ON network.
  • Low blocking probability.
  • Runs every hour.
  • Exchange of information overhead between cells every hour.
  • Low power saving 7:00 pm to 11:00 pm.
Micro DTX scheme [60]
  • Power saving up to 61% compared to a cell without any DTX.
  • No need to power off the whole BS.
  • Uses MBSFN subframes for power saving.
  • Creates empty transmission intervals during which PA can be deactivated.
  • Longer sleep mode increases delays; 10–20 s in going back to active mode.
  • Increased number of MBSFN subframes decreases the capacity and bandwidth.
  • In LTE rel‐8, information could change at the system broadcast channel only once in every 6 min.
Enhanced DTX scheme [60]
  • Power saving up to 89% compared to a cell without any DTX.
  • Only synchronization and other secondary signals transmitted.
  • Increased number of MBSFN subframes decreases the capacity and bandwidth
  • In LTE rel‐8, information could change at the system broadcast channel only once every 6 min.
Energy efficient bandwidth expansion scheme [96]
  • Saves power up to 44%
  • Effective for a lightly loaded network.
  • Reduces the actual capacity of bandwidth.
  • As traffic load increases, bandwidth decreases.
Centralized Algorithm [73]
  • Uses the load information scope from the entire network.
  • More effective with a lower number of users.
  • Lower transition cost with low bandwidth requirements
  • Higher worst‐case complexity due to binary heaps.
TCoM [85]
  • Provide ES up to 26% compared to an always ON System.
  • Deactivation of RBs is a very effective ES technique.
  • Ineffective ES at cell edges.
  • Suffers from capacity limitation.
EE Link Adaptation Scheme [78]
  • More effective for the UEs closer to the BS.
  • Limited reduction.
  • Increased feedback overhead.
BEM [83]
  • Significantly reduced power consumption at BS in low load cells.
  • Distributes users only from the overlapping area between two cells, thus reducing overheads.
  • Allows the use of lower order modulation schemes, which consume less power.
  • Not suitable for highly loaded cells.
  • ES in trade‐off with more bandwidth used.
  • Distributes users to those cells that are already desirable for ES mode, thus reduce power saving opportunities in the overall network.
CC Based EE Scheme [65]
  • Supports both real and non‐real time traffic simultaneously.
  • Reduces power consumption by 50% compared to an always ON CC Network.
  • Only considers two CCs.
  • Session blocking increases the delay in a high traffic period.
Energy Efficient BS deployment [87]
  • Provides static hardware based ES.
  • Scheme does not provide further ES opportunities once BSs have deployed.
Power Aware allocation of MBSFN subframes [61]
  • Uses two power saving concepts.
  • Deactivates unused subframes.
  • Allocates RBs as much as required depending on load.
  • Increased delays.
  • Only a few subframes can be switched off because control signals require capacity in a few subframes.
Coverage Based Scheme [72]
  • Significantly reduces power consumption.
  • Recommended for a lightly loaded network.
  • Only one partition of BS could be expended rather than full BS coverage area.
  • Execution of multiple algorithms increases processing computation.
  • Challenging for a BS to use one sector to provide coverage to the full area of neighbour BSs.
  • Load information overheads.
Sector Based Scheme [71]
  • SON based ES scheme.
  • Distributed in nature; each BS provides ES without communicating with other BSs.
  • Divides coverage in different numbers of sector depending on load.
  • Additional processing computation for sectorization of BS coverage.
  • Even existence of users in each sector reduces ES opportunities for BS.
  • Challenging for BSs to manage varying sectors in their coverage.
Relay Based Scheme [77]
  • Significantly reduces power consumption from around 15–20%.
  • UEs can freely access recourses of two different operators.
  • Does not require load information exchange through an X2 interface.
  • Difficult for two operators to work together
  • Allocation of resources to another operator’s UEs may cause capacity limitation for their own associated UEs.
CRN Based Schemes [91]
  • Turns off only unused resources during idle time frames.
  • Secondary users wait until sleep mode has completed thus, resulting in delay.
REHO Scheme [97]
  • Early handover helps resources to become free earlier and are turned off for ES.
  • Increased radio link failure due to early handover.

Since each BS is required to exchange load information every hour with other neighbour cells, this results in an increased overhead in the network. DTX is one of the most interesting ES schemes. The main advantage of DTX is that it targets operational ES where there is no need of turning off whole BS and only unused RBs are switched off [60, 61]. On the other hand, the main disadvantage of DTX is the long sleep mode of unused RBs that increases delay time required by RBs to go back to the active mode. Distributed schemes also contribute in ES in LTE networks by effectively migrating UEs to the neighbour cells [72, 73]. In these types of schemes, BSs keep exchanging load information with each other resulting in an increased traffic information overhead. Bandwidth expansion is also used to achieve 44% of ES in lightly loaded networks. However, allocation of extra RBs results in reduced available capacity of the BS and thus is not very effective during the peak hour time period [96]. A combination of load balancing with bandwidth expansion is also used to reduce power consumption in the network [83]. However, this scheme could migrate UEs to those cells that are already desirable for ES mode, thus reducing ES opportunities in the overall network. The centralized schemes also provide ES but suffer from high traffic load similar to the distributed scheme. The TCoM scheme provides 26% ES by cutting down control channel signalling [85]. The main idea is similar to bandwidth expansion; however, it reduces the control channel overhead by transmitting two RBs jointly to a single user. TCoM suffers from the drawback of being not effective at cell edges and also requires a reduced overall system capacity. The EE link adaptation scheme is only effective for UEs located closer to the BS and saves 9.4% of energy while increasing the feedback overhead in the network [78]. The carrier aggregation approach is also used for ES reducing power consumption by 50% compared to an always‐on network [65]. One of the disadvantages of this scheme is session blocking, which may increase the delay during high traffic time period. Coverage expansion is also used as a means to realize ES in the LTE networks [72]. It is, however, very complicated for the BS to make partitions to expand their transmission power and provide coverage to full neighbour BSs using one partition. On the other hand, execution of two algorithms to implement this scheme also increases overall computation overhead. Division of the BS in different sectors and turning off unused or free sectors is also employed to attain ES [71]. This scheme, however, only works for a completely free sector and existence of even a single user would not allow the BS to turn off that sector. A shared relay ES scheme based on the idea of sharing resources of two different operators is proposed in [77]. However, it is very difficult for two different operators to work and integrate their operations under a shared environment.

Table 5.3 summarizes the performance of the discussed ES schemes in relation to other QoS issues. Figure 5.17 presents the ES percentage achieved by different ES schemes discussed before. It can be observed that dynamic distance aware scheme is the most EE technique of other ES schemes.

Table 5.3 QoS factors involved in ES schemes.

ES Scheme QoS Issues
Lightly loaded Heavily loaded Reduced capacity Increased
delay		 Increased overhead
Distance Aware [68] ✓ X X X ✓
Dynamic Distance aware [69] ✓ X X X ✓
Micro DTX scheme [60] ✓ ✓ X ✓ X
Bandwidth expansion [83] ✓ X ✓ X X
Centralized Algorithm [73] ✓ X X X ✓
TCoM [85] ✓ ✓ X X X
EE Link Adaptation [78] ✓ ✓ ✓ X ✓
Component Carrier [65] ✓ ✓ ✓ X X
EE BS deployment [87] ✓ ✓ X X X
Power Aware MBSFN [61] ✓ ✓ ✓ ✓ X
Coverage Based Scheme [72] ✓ X X ✓ ✓
Sector Based Scheme [71] ✓ X X ✓ X
Relay Based Scheme [77] ✓ ✓ X ✓ ✓
CRN Based Scheme [91] ✓ X X ✓ X
REHO Scheme [97] ✓ ✓ ✓ X X
Bar graph depicting the percentage of energy saved in each ES scheme, 8 vertical bars arranged in ascending order marked from 1 to 8 representing EE Link adaptation ES scheme, relay based ES scheme, TCoM ES scheme, etc.

Figure 5.17 Percentage of energy saved in each ES scheme.

5.8 Open Research Issues

Comparative study of various existing ES schemes has shown that most of these schemes are only effective for lightly loaded networks and energy is not saved during highly loaded network conditions. DTX based schemes affected from delay that occur for RBs to come back in active mode [60]. Further research work is needed to reduce these delays. Reduced delay could have significant effect on overall performance of the system. The distance aware and bandwidth expansion based schemes fail to reduce power consumption during peak traffic hours. Therefore, these schemes could be further explored to provide enhanced ES during highly loaded traffic [68, 69]. Bandwidth expansion schemes could work more effectively in a balanced network. Therefore, load balancing could be further exploited with bandwidth expansion [83, 96]. On the other hand, centralized and distributed schemes exchange load information between the entire BS, which increases load information overheads in a network and reduces system efficiency [72, 73]. Means should be devised to reduce the load information overhead. Similarly, a link adaptation based ES scheme also suffers from overheads produced by energy consumption feedback sent to the BS [78]. Feedback overhead reduction could be exploited for an improved ES in the LTE networks. EE BS deployment provision could be integrated with any other dynamic ES based schemes for enhanced EE systems [87]. An MBSFN based ES scheme suffers from control signals that basically reduce the opportunities for turning off the unused RBs [61]. ES through control signals could be further explored for enhanced MBSFN based ES. In other words, a few aspects of both the TCoM and MBSFN schemes can be taken into account to develop a hybrid ES scheme that may provide a better EE system compared to systems using TCoM and MBSFN schemes on an individual basis [61, 87]. Table 5.4 presents open research areas for ES in LTE networks.

Table 5.4 Open research issues.

ES Scheme Open Research Areas
Distance Aware [68] ES can be extended for 24 h including daytime.
Dynamic Distance aware [69] Load information overhead could be reduced for enhanced system performance, whereas ES could be extended to include daytime period.
Micro DTX scheme [60] Delay could be reduced or controlled for better performance.
Enhanced DTX scheme [60] Capacity limitation could be explored as an open research issue.
Centralized Algorithm [73] Complex system due to binary heaps could be explored for better performing ES scheme.
TCoM [85] ES could be further improved at cell edges for enhanced system performance.
EE Link Adaptation Scheme [78] Feedback overhead could be considered as an open research issue.
BEM [83] Further enhancement can be done offering ES during peak load hours.
Component Carrier Based EE Scheme [65] More CCs can be considered in future research work.
Energy Efficient BS deployment [87] ES could be extended towards dynamic part of an enhanced system.
Power Aware allocation of MBSFN subframes [61] Capacity limitations can be studied as an open research issue.

Though various schemes have been discussed that help to achieve ES, thereby improving energy efficiency, significant lessons, however, have been learnt. DTX provides significant ES; however, it does not consider any delay related issues that may occur due to the longer sleep mode from a 10–20 ms cycle. Similarly, the distance aware switching off of BSs is effective for lightly loaded networks; however, these schemes are not at all suitable for heavily loaded networks. Since it is not possible to turn off a BS during peak load hours, therefore these schemes fail to save energy. Bandwidth expansion schemes allocate extra RBs to UEs that reduce overall available capacity of the system. Accordingly, BEM does not provide ES opportunities during peak hour traffic. Further exploring the literature review, TCoM considered constant SINR and provides ES through reduced control channels overheads, yet RBs still consume power during idle mode and TCoM loses ES opportunities. Importantly, centralized schemes have also proven to be effective for lightly loaded networks only; these schemes fail to provide ES in highly loaded networks. Link adaptation schemes reduce power consumption by only 9.4%, while it considerably increases overhead of feedback sent back to BS for ES purpose. The carrier aggregation based ES scheme combines two carrier components whereas it suffers from session blocking resulting in increased delay during high traffic periods. Most existing ES proposals mostly work in lightly loaded networks while they do not provide adequate and considerable ES during peak hours. Indeed, plenty of room and a need for additional research work exists that could be carried out to provide ES during peak traffic time periods too. The mentioned drawbacks and learnt lessons can indeed be carefully employed and guide future researchers to conduct and develop novel robust EE schemes in the light of discussed open research issues (Table 5.4).

5.9 Summary

Due to increased global warming and worldwide climate change, energy consumption has become a major hurdle. The ICT contributes approximately 2% to global warming, while a major part is attributed to telecommunications in ICT. In cellular networks, energy consumption is affected by growing mobile users and their data requirements. Moreover, further deployment of additional and enhanced BSs to fulfil ever growing UEs also adds in ICT contribution. Therefore, the increasing trend for energy consumption has become a major challenge for vendors thus affecting both economic and environmental aspects. The rapid increase in energy consumption not only increases OPEX but also effects climate change. Research work has proven that BSs in LTE networks consume a lot of dynamic power during the idle state, which could be saved by appropriate ES schemes. The reduced power consumption enhances the LTE system performance through cutting down OPEX and carbon emissions, thus also helping vendors to have a high profile in green communication. This chapter has provided a detailed discussion of existing ES schemes developed for LTE networks. Critical analysis of the schemes has been presented before open research issues were discussed. Finally, the authors’ novel REHO ES scheme was also briefly described that takes into account challenges and builds itself on research issues. The chapter is a comprehensive account of the existing ES schemes for LTE networks and can help researchers to understand the current state‐of‐the‐art, open research issues to come up with innovative solutions resulting in optimized system performance.

References

  1. 1 M. A. Garrett, “Radio astronomy transformed: Aperture arrays – Past, present and future,” AFRICON, 2013, Pointe‐Aux‐Piments, 2013, pp. 1–5.
  2. 2 Q. Fan, H. Lu, P. Hong and Z. Zhu, “Throughput‐power tradeoff association for user equipment in WLAN/cellular integrated networks,” in IEEE Transactions on Vehicular Technology, vol. 66, no. 4, pp. 3462–3474, 2017.
  3. 3 Y. W. Blankenship, “Achieving high capacity with small cells in LTE‐A,” Communication, Control, and Computing (Allerton), 2012 50th Annual Allerton Conference on, Monticello, IL, 2012, pp. 1680–1687.
  4. 4 M. Agiwal, A. Roy and N. Saxena, “Next generation 5G wireless networks: A comprehensive survey,” in IEEE Communications Surveys & Tutorials, vol. 18, no. 3, pp. 1617–1655, third quarter 2016.
  5. 5 A. Damnjanovic, J. Montojo, and Y. Wei et al., “A survey on 3GPP heterogeneous networks,” in IEEE Wireless Communications, vol. 18, no. 3, pp. 10–21, June 2011.
  6. 6 F. Adachi, “Keynote talk #1: Wireless evolution and challenges for 5G wireless networks,” Information and Computer Science (NICS), 2015 2nd National Foundation for Science and Technology Development Conference on, Ho Chi Minh City, 2015, pp. xxi–xxii.
  7. 7 N. Chandran and M. C. Valenti, “Three generations of cellular wireless systems,” in IEEE Potentials, vol. 20, no. 1, pp. 32–35, Feb/Mar 2001.
  8. 8 B. Esmailpour, S. Salehi and N. Safavi, “Quality of service differentiation measurements in 4G networks,” Wireless Telecommunications Symposium (WTS), 2013, Phoenix, AZ, 2013, pp. 1–5.
  9. 9 Y. Tian, A. Nix and M. Beach, “4G femtocell LTE base station with diversity and adaptive antenna techniques,” Wireless Communications, Networking and Mobile Computing (WiCOM 2014), 10th International Conference on, Beijing, 2014, pp. 216–221.
  10. 10 3GPP, “Evolved Universal Terrestrial Radio Access (E‐UTRA): Base Station (BS) radio transmission and reception,” TS 36.104, V11.9.0.
  11. 11 M. Nasimi, F. Hashim and C. K. Ng, “Characterizing energy efficiency for heterogeneous cellular networks,” IEEE Student Conference on Research and Development (SCOReD), 2012, Pulau Pinang, 2012, pp. 198–202.
  12. 12 Shayea, M. Ismail and R. Nordin, “Advanced handover techniques in LTE‐advanced system,” International Conference on Computer and Communication Engineering (ICCCE), 2012, Kuala Lumpur, 2012, pp. 74–79.
  13. 13 U. Dampage and C. B. Wavegedara, “A low‐latency and energy efficient forward handover scheme for LTE‐femtocell networks,” 2013 IEEE 8th International Conference on Industrial and Information Systems, Peradeniya, 2013, pp. 53–58.
  14. 14 P. Skocir, D. Katusic, I. Novotni, I. Bojic and G. Jezic, “Data rate fluctuations from user perspective in 4G mobile networks,” 22nd International Conference on Software, Telecommunications and Computer Networks (SoftCOM), 2014, Split, 2014, pp. 180–185.
  15. 15 S. Abeta, “Toward LTE commercial launch and future plan for LTE enhancements (LTE‐Advanced),” IEEE International Conference on Communication Systems (ICCS), 2010, Singapore, 2010, pp. 146–150.
  16. 16 M. F. L. Abdullah and A. Z. Yonis, “Performance of LTE Release 8 and Release 10 in wireless communications,” Cyber Security, Cyber Warfare and Digital Forensic (CyberSec), 2012 International Conference on, Kuala Lumpur, 2012, pp. 236–241.
  17. 17 N. Mehra, and A. Noliya, “Performance analysis of OFDMA, MIMO and SC‐FDMA technology in 4G LTE networks,” 2016 6th International Conference – Cloud System and Big Data Engineering (Confluence), Noida, 2016, pp. 554–558.
  18. 18 N. Takaharu, “LTE and LTE‐advanced: Radio technology aspects for mobile communications,” General Assembly and Scientific Symposium, 2011 30th URSI, Istanbul, 2011, pp. 1–4.
  19. 19 H. Lee, S. Vahid and K. Moessner, “A Survey of Radio Resource Management for Spectrum Aggregation in LTE‐Advanced,” in IEEE Communications Surveys & Tutorials, vol. 16, no. 2, pp. 745–760, Second Quarter 2014.
  20. 20 N. Becker, A. Rizk and M. Fidler, “A measurement study on the application‐level performance of LTE,” Networking Conference, 2014 IFIP, Trondheim, 2014, pp. 1–9.
  21. 21 S. B. Manir, M. M. Rahman and T. Ahmed, “Comparison between FDD and TDD frame structure in SC‐FDMA,” International Conference on Informatics, Electronics & Vision (ICIEV), 2012, Dhaka, 2012, pp. 795–799.
  22. 22 L. Wan, M. Zhou and R. Wen, “Evolving LTE with flexible duplex,” 2013 IEEE Globecom Workshops (GC Wkshps), Atlanta, GA, 2013, pp. 49–54.
  23. 23 R. Ratasuk, A. Ghosh, Weimin Xiao, R. Love, R. Nory and B. Classon, “TDD design for UMTS long‐term evolution,” 2008 IEEE 19th International Symposium on Personal, Indoor and Mobile Radio Communications, Cannes, 2008, pp. 1–5.
  24. 24 R. Zheng, X. Zhang, X. Li, Q. Pan, Y. Fang and D. Yang, “Performance evaluation on the coexistence scenario of two 3GPP LTE systems,” IEEE 70th Vehicular Technology Conference Fall (VTC 2009‐Fall), 2009, Anchorage, AK, 2009, pp. 1–6.
  25. 25 S. S. Prasad, C. K. Shukla and R. F. Chisab, “Performance analysis of OFDMA in LTE,” 2012 Third International Conference on Computing Communication & Networking Technologies (ICCCNT), Coimbatore, 2012, pp. 1–7.
  26. 26 L. A. M. R. de Temino, G. Berardinelli, S. Frattasi, K. Pajukoski and P. Mogensen, “Single‐user MIMO for LTE‐A Uplink: Performance evaluation of OFDMA vs. SC‐FDMA,” 2009 IEEE Radio and Wireless Symposium, San Diego, CA, 2009, pp. 304–307.
  27. 27 3GPP, “Evolved Universal Terrestrial Radio Access (E‐UTRA) and Evolved Universal Terrestrial Radio Access Networks (E UTRAN): Overall description,” V10.4.0., TS 36.300.
  28. 28 3GPP, “Technical Specification Group Radio Access Network; Further advancements for E‐UTRA – LTE‐Advanced feasibility studies in RAN WG4,” V9.0.0., TR 36.815.
  29. 29 3GPP, “General Packet Radio Service (GPRS) enhancements for Evolved Universal Terrestrial Radio Access Network (E‐UTRAN) access,” 3rd Generation Partnership Project (3GPP), TS 23.401, Jun. 2011.
  30. 30 S. B. H. Said, M. R. Sama, K. Guillouard, et al., “New control plane in 3GPP LTE/EPC architecture for on‐demand connectivity service,” IEEE 2nd International Conference on Cloud Networking (CloudNet), 2013, San Francisco, CA, 2013, pp. 205–209.
  31. 31 C. Cox, An Introduction to LTE: LTE, LTE‐Advanced, SAE and 4G Mobile Communications, Chichester, UK: John Wiley & Sons, Ltd, 2012, pp. 21–28.
  32. 32 K. P. Makhecha and K. H. Wandra, “4G wireless networks: Opportunities and challenges,” 2009 Annual IEEE India Conference, Gujarat, 2009, pp. 1–4.
  33. 33 A. Pande, V. Ahuja, R. Sivaraj, E. Baik and P. Mohapatra, “Video delivery challenges and opportunities in 4G networks,” IEEE MultiMedia, vol. 20, no. 3, pp. 88–94, July–Sept. 2013.
  34. 34 M. Iwamura, K. Etemad, M. H. Fong, R. Nory and R. Love, “Carrier aggregation framework in 3GPP LTE‐advanced [WiMAX/LTE Update],” in IEEE Communications Magazine, vol. 48, no. 8, pp. 60–67, August 2010.
  35. 35 Y. Yuan, S. Wu, J. Yang, F. Bi, S. Xia and G. Li, “Relay backhaul subframe allocation in LTE‐Advanced for TDD,” 5th International ICST Conference on Communications and Networking in China (CHINACOM), 2010, Beijing, 2010, pp. 1–5.
  36. 36 S. Ahmadi, LTE‐advanced: A practical systems approach to understanding 3GPP LTE Release 10 and 11 Radio Access Technologies, Waltham, MA, USA, Elsevier Inc., 2014, pp. 61–65.
  37. 37 G. Cili, H. Yanikomeroglu and F. R. Yu, “Cell switch off technique combined with coordinated multi‐point (CoMP) transmission for energy efficiency in beyond‐LTE cellular networks,” 2012 IEEE International Conference on Communications (ICC), Ottawa, ON, 2012, pp. 5931–5935.
  38. 38 M. Pickavet, W. Veerecken, S. Demeyer, et al., “Worldwide energy needs for ICT: The rise of power‐aware networking,” 2008 2nd International Symposium on Advanced Networks and Telecommunication Systems, Mumbai, 2008, pp. 1–3.
  39. 39 Y. L. Chung, “Energy‐saving transmission for green macrocell–small cell systems: A system‐level perspective,” IEEE Systems Journal, vol. 11, no. 2, pp. 706–716, 2017.
  40. 40 E. Oh, B. Krishnamachari, X. Liu and Z. Niu, “Toward dynamic energy‐efficient operation of cellular network infrastructure,” in IEEE Communications Magazine, vol. 49, no. 6, pp. 56–61, June 2011.
  41. 41 R. Maihaniemi, “ICT Getting Green,” 4th International Conference on Telecommunication – Energy Special Conference (TELESCON), 2009, Vienna, Austria, 2009, pp. 1–6.
  42. 42 H. O. Scheck, “ICT & wireless networks and their impact on global warming,” European Wireless Conference (EW), 2010, Lucca, 2010, pp. 911–915.
  43. 43 Y. Chen, S. Zhang, S. Xu and G. Y. Li, “Fundamental trade‐offs on green wireless networks,” in IEEE Communications Magazine, vol. 49, no. 6, pp. 30–37, June 2011.
  44. 44 G. Fettweis and E. Zimmermann, “ICT energy consumption – Trends and challenges,” in Proc. 11th Int. Symp. WPMC, vol. 2, no. 4, Sep. 2008.
  45. 45 M. Griffiths, “ICT and CO2 Emissions,” Parliamentary Office Sci. Technol., London, Postnote No. 319, Dec. 2008. [Online]. Available: www.parliament.uk/documents/post/postpn319.pdf (accessed December 2017).
  46. 46 T. M. Knoll, “A combined CAPEX and OPEX cost model for LTE networks,” Telecommunications Network Strategy and Planning Symposium (Networks), 2014 16th International, Funchal, 2014, pp. 1–6.
  47. 47 J. Moysen and L. Giupponi, “Self coordination among SON functions in LTE heterogeneous networks,” 2015 IEEE 81st Vehicular Technology Conference (VTC Spring), Glasgow, 2015, pp. 1–6.
  48. 48 T. M. Knoll, “Life‐cycle cost modelling for NFV/SDN based mobile networks,” Telecommunication, Media and Internet Techno‐Economics (CTTE), 2015 Conference of, Munich, 2015, pp. 1–8.
  49. 49 N. Zhang and H. Hämmäinen, “Cost efficiency of SDN in LTE‐based mobile networks: Case Finland,” International Conference and Workshops on Networked Systems (NetSys), 2015, Cottbus, 2015, pp. 1–5.
  50. 50 S. M. Azzam and T. Elshabrawy, “Re‐dimensioning number of active eNodeBs for green LTE networks using genetic algorithms,” Proceedings of European Wireless 2015; 21th European Wireless Conference, Budapest, Hungary, 2015, pp. 1–6.
  51. 51 O. Arnold, F. Richter, G. Fettweis and O. Blume, “Power consumption modelling of different base station types in heterogeneous cellular networks,” 2010 Future Network & Mobile Summit, Florence, 2010, pp. 1–8.
  52. 52 K. Hiltunen, “Utilizing eNodeB sleep mode to improve the energy‐efficiency of dense LTE networks,” 2013 IEEE 24th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), London, 2013, pp. 3249–3253.
  53. 53 T. Chen, H. Zhang, Z. Zhao and X. Chen, “Towards green wireless access networks,” 5th International ICST Conference on Communications and Networking in China (CHINACOM), 2010, Beijing, 2010, pp. 1–6.
  54. 54 S. Srikanth, P. A. Murugesa Pandian and X. Fernando, “Orthogonal frequency division multiple access in WiMAX and LTE: a comparison,” in IEEE Communications Magazine, vol. 50, no. 9, pp. 153–161, September 2011.
  55. 55 S. Janaaththanan, C. Kasparis and B. G. Evans, “Comparison of SC‐FDMA and HSUPA in the return‐link of evolved S‐UMTS architecture,” Satellite and Space Communications, 2007. IWSSC ’07. International Workshop on, Salzburg, 2007, pp. 56–60.
  56. 56 M. Salem, A. Adinoyi, H. Yanikomeroglu and Y. D. Kim, “Radio resource management in OFDMA‐based cellular networks enhanced with fixed and nomadic relays,” 2010 IEEE Wireless Communication and Networking Conference, Sydney, NSW, 2010, pp. 1–6.
  57. 57 M. M. Matalgah, B. Paudel and O. M. Hammouri, “Cross‐layer resource allocation approach in OFDMA systems with multi‐class QoS services and users queue status,” 2013 IEEE Global Communications Conference (GLOBECOM), Atlanta, GA, 2013, pp. 1385–1390.
  58. 58 M. Deruyck, E. Tanghe, W. Joseph and L. Martens, “Modelling the energy efficiency of microcell base stations,” in 1st International Conference on Smart Grids, Green Communications and IT Energy‐Aware Technologies, 2012, p. 1–6.
  59. 59 W. Tomaselli, D. Sabella, V. Palestini and V. Bernasconi, “Energy efficiency performances of selective switch OFF algorithm in LTE mobile networks,” 2013 IEEE 24th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), London, 2013, pp. 3254–3258
  60. 60 P. Frenger, P. Moberg, J. Malmodin, Y. Jading and I. Godor, “Reducing energy consumption in LTE with cell DTX,” 2011 IEEE 73rd Vehicular Technology Conference (VTC Spring), Yokohama, 2011, pp. 1–5.
  61. 61 A. Migliorini, G. Stea, M. Caretti and D. Sabella, “Power‐aware allocation of MBSFN subframes using discontinuous cell transmission in LTE systems,” Vehicular Technology Conference (VTC Fall), 2013 IEEE 78th, Las Vegas, NV, 2013, pp. 1–5.
  62. 62 S. Herrería‐Alonso, M. Rodríguez‐Pérez, M. Fernández‐Veiga and C. López‐García, “Adaptive DRX scheme to improve energy efficiency in LTE networks with bounded delay,” in IEEE Journal on Selected Areas in Communications, vol. 33, no. 12, pp. 2963–2973, Dec. 2015.
  63. 63 K. Davaslioglu, C. C. Coskun and E. Ayanoglu, “Energy‐efficient resource allocation for fractional frequency reuse in heterogeneous networks,” in IEEE Transactions on Wireless Communications, vol. 14, no. 10, pp. 5484–5497, Oct. 2015.
  64. 64 X. Xiao, X. Tao and J. Lu, “Energy‐efficient resource allocation in LTE‐Based MIMO‐OFDMA systems with user rate constraints,” in IEEE Transactions on Vehicular Technology, vol. 64, no. 1, pp. 185–197, Jan. 2015.
  65. 65 A. T. Tung, Y. L. Chung and Z. Tsai, “An efficient power‐saving downlink transmission scheme in OFDM‐based multiple component carrier systems,” Advanced Communication Technology (ICACT), 2012 14th International Conference on, PyeongChang, 2012, pp. 116–120.
  66. 66 K. Sundaresan and S. Rangarajan, “Energy efficient carrier aggregation algorithms for next generation cellular networks,” 2013 21st IEEE International Conference on Network Protocols (ICNP), Goettingen, 2013, pp. 1–10
  67. 67 A. Liu, K. Zheng, W. Xiang and H. Zhao, “Design and performance analysis of an energy‐efficient uplink carrier aggregation scheme,” in IEEE Journal on Selected Areas in Communications, vol. 32, no. 2, pp. 197–207, February 2014.
  68. 68 A. Bousia, A. Antonopoulos, L. Alonso and C. Verikoukis, ““Green” distance‐aware base station sleeping algorithm in LTE‐Advanced,” 2012 IEEE International Conference on Communications (ICC), Ottawa, ON, 2012, pp. 1347–1351.
  69. 69 A. Bousia, E. Kartsakli, L. Alonso and C. Verikoukis, “Dynamic energy efficient distance‐aware base station switch on/off scheme for LTE‐advanced,” Global Communications Conference (GLOBECOM), 2012 IEEE, Anaheim, CA, 2012, pp. 1532–1537.
  70. 70 L. Li, Y. Zhang, B. Fan and H. Tian, “Mobility‐aware load balancing scheme in hybrid VLC‐LTE networks,” in IEEE Communications Letters, vol. 20, no. 11, pp. 2276–2279, Nov. 2016.
  71. 71 M. F. Hossain, K. S. Munasinghe and A. Jamalipour, “Toward self‐organizing sectorization of LTE eNBs for energy efficient network operation under QoS constraints,” 2013 IEEE Wireless Communications and Networking Conference (WCNC), Shanghai, 2013, pp. 1279–1284.
  72. 72 K. Samdanis, D. Kutscher and M. Brunner, “Self‐organized energy efficient cellular networks,” IEEE 21st International Symposium on Personal Indoor and Mobile Radio Communications (PIMRC), 26–30 Sept. 2010, pp. 1665, 1670.
  73. 73 K. Samdanis, T. Taleb, D. Kutscher and M. Brunner, “Self organized network management functions for energy efficient cellular urban infrastructures” Mob. Netw. Appl., vol. 17, pp. 119–131, 2012.
  74. 74 S. A. Ahmad and D. Datla, “Distributed power allocations in heterogeneous networks with dual connectivity using backhaul state information,” IEEE Transactions on Wireless Communications, vol. 14, no. 8, pp. 4574–4581, Aug. 2015.
  75. 75 H. H. Nguyen and W. J. Hwang, “Distributed scheduling and discrete power control for energy efficiency in multi‐cell networks,” IEEE Communications Letters, vol. 19, no. 12, pp. 2198–2201, Dec. 2015.
  76. 76 S. A. Ahmad and D. Datla, “Distributed power allocations in heterogeneous networks with dual connectivity using backhaul state information,” IEEE Transactions on Wireless Communications, vol. 14, no. 8, pp. 4574–4581, Aug. 2015.
  77. 77 V. Monteiro, T. Ramrekha, D. Yang, J. Rodriguez, S. Mumtaz and C. Politis, “An energy efficient proposal in shared relay‐based LTE network,” Wireless Communication Systems (ISWCS 2013), Proceedings of the Tenth International Symposium on, Ilmenau, Germany, 2013, pp. 1–5.
  78. 78 A. Li, S. Jin, F. Zheng, X. Gao and X. Wang, “Energy efficient link adaptation for downlink transmission of LTE/LTE‐A systems,” Vehicular Technology Conference (VTC Fall), 2013 IEEE 78th, Las Vegas, NV, 2013, pp. 1–5.
  79. 79 D. D. Ling, Z. Lu, W. Zheng, X. Wen and Y. Ju, “Energy efficient cross‐layer resource allocation scheme based on potential games in LTE‐A,” 15th International Symposium on Wireless Personal Multimedia Communications (WPMC), 2012, Taipei, pp. 623–627.
  80. 80 Z. Zhou, M. Dong, K. Ota, G. Wang and L. T. Yang, “Energy‐efficient resource allocation for D2D communications underlaying cloud‐RAN‐based LTE‐A networks,” IEEE Internet of Things Journal, vol. 3, no. 3, pp. 428–438, June 2016.
  81. 81 S. Rostami, K. Arshad and P. Rapajic, “Energy‐efficient resource allocation for LTE‐A networks,” IEEE Communications Letters, vol. 20, no. 7, pp. 1429–1432, July 2016.
  82. 82 A. Z. Kaddour, E. Vivier, L. Mroueh, M. Pischella and P. Martins, “Green opportunistic and efficient resource block allocation algorithm for LTE uplink networks,” IEEE Transactions on Vehicular Technology, vol. 64, no. 10, pp. 4537–4550, Oct. 2015.
  83. 83 Y. Li, W. Liu, B. Cao and M. Li, “Green resource allocation in LTE system for unbalanced low load networks,” 2012 IEEE 23rd International Symposium on Personal, Indoor and Mobile Radio Communications – (PIMRC), Sydney, NSW, 2012, pp. 1009–1014.
  84. 84 S. Almowuena, M. M. Rahman, C. H. Hsu, A. A. Hassan and M. Hefeeda, “Energy‐aware and bandwidth‐efficient hybrid video streaming over mobile networks,” IEEE Transactions on Multimedia, vol. 18, no. 1, pp. 102–115, Jan. 2016.
  85. 85 S. Videv, H. Haas, J. S. Thompson and P. M. Grant, “Energy efficient resource allocation in wireless systems with control channel overhead,” Wireless Communications and Networking Conference Workshops (WCNCW), 2012 IEEE, Paris, 2012, pp. 64–68.
  86. 86 A. Han and S. Armour, “Energy efficient radio resource management strategies for green radio,” IET Communications, vol. 5, no. 18, pp. 2629–2639, Dec. 16 2011.
  87. 87 C. Coskun and E. Ayanoglu, “Energy‐efficient base station deployment in heterogeneous networks,” IEEE Wireless Communications Letters, vol. 3, no. 6, pp. 593–596, Dec. 2014.
  88. 88 P. Van, B. P. Rimal, S. Andreev, T. Tirronen and M. Maier, “Machine‐to‐machine communications over WiFi enhanced LTE networks: A power‐saving framework and end‐to‐end performance,” Journal of Lightwave Technology, vol. 34, no. 4, pp. 1062–1071, Feb.15, 15 2016.
  89. 89 S. Mumtaz, K. M. Saidul Huq, J. Rodriguez and V. Frascolla, “Energy‐efficient interference management in LTE‐D2D communication,” IET Signal Processing, vol. 10, no. 3, pp. 197–202, 5 2016.
  90. 90 J. Liu, N. Kato, J. Ma and N. Kadowaki, “Device‐to‐device communication in LTE‐advanced networks: a survey,” IEEE Communications Surveys & Tutorials, vol. 17, no. 4, pp. 1923–1940, Fourthquarter 2015.
  91. 91 S. Jin, X. Ma and W. Yue, “Energy‐saving strategy for green cognitive radio networks with an LTE‐advanced structure,” Journal of Communications and Networks, vol. 18, no. 4, pp. 610–618, Aug. 2016.
  92. 92 F. Zheng, W. Li, L. Meng, P. Yu and L. Peng, “Distributed energy saving mechanism based on CoMP in LTE‐A system,” China Communications, vol. 13, no. 7, pp. 39–47, July 2016.
  93. 93 V. J. Kotagi, R. Thakur, S. Mishra and C. S. R. Murthy, “Breathe to Save Energy: Assigning downlink transmit power and resource blocks to LTE enabled IoT networks,” IEEE Communications Letters, vol. 20, no. 8, pp. 1607–1610, Aug. 2016.
  94. 94 L. You, L. Lei and D. Yuan, “Optimizing power and user association for energy saving in load‐coupled cooperative LTE,” 2016 IEEE International Conference on Communications (ICC), Kuala Lumpur, 2016, pp. 1–6.
  95. 95 A. Kalogridis and O. Georgiou, “CreD2D: A credit‐driven self‐evolving D2D towards LTE HetNet energy saving,” 2016 IEEE International Conference on Communications Workshops (ICC), Kuala Lumpur, 2016, pp. 742–748.
  96. 96 S. Videv and H. Haas, “Energy‐efficient scheduling and bandwidth‐energy efficiency trade‐off with low load,” 2011 IEEE International Conference on Communications (ICC), Kyoto, 2011, pp. 1–5.
  97. 97 K. Kanwal and G. A. Safdar, “Reduced early handover for energy saving in LTE Networks,” IEEE Communications Letters, vol. 20, no. 1, pp. 153–156, Jan. 2016.
  98. 98 G. A. Safdar, M. Ur‐Rehman, M. Muhammad, M. A. Imran and R. Tafazolli, “Interference mitigation in D2D Communication underlaying LTE‐A Network,” IEEE Access, vol. 4, pp. 7967–7987, 2016.
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset