Point-to-Point Protocol Over ATM (PPPoA)

This section describes the encapsulations, data flow, IP addressing considerations, and design considerations required for implementing an end-to-end ADSL architecture using PPPoA. PPPoA has created high interest and is gaining service provider popularity. It will form a larger portion of future ADSL deployments.

PPPoA Overview

PPP is defined in RFCs 1331 and 1661 as a standard method of encapsulating different types of higher-layer protocol datagrams across point-to-point connections. It extends the high-level data link control (HDLC) packet structure with a 16-bit protocol identifier that contains information about the packet's contents.

PPP is defined for use specifically with AAL5 in RFC 2364. PPPoA operates in either of two modes—LLC/SNAP or Virtual Circuit-multiplexing (VC-mux).

In the PPPoA model, as shown in Figure 3-9, the ATU-R encapsulates user data (PDUs) of various types in PPP frames. With PPPoA, the PPP session is between the ATU-R and the aggregation router; the PPP link is terminated at the provider's aggregation router.

This Layer 3 aggregation router typically refers to a RADIUS server used to authenticate and authorize the user, although this can be done within the L3 device itself. Host routes are installed during PPP negotiation on both the ATU-R and the aggregator, and it is over these routes that the packets are forwarded. The aggregation router can be an IP DSL Switch or a universal concentrator if Layer 2 DSLAMs are used instead of IP DSL Switches. Subinterfaces on the aggregator are configured for PPP and must be bound to a virtual template interface.

Depending on the capabilities of the particular ATU-R model and software, the ATU-R can receive its IP address from the aggregation router via IPCP. It can be configured as the DHCP server for the subscriber PCs, and it can be configured to perform NAT/PAT if private addressing is used for the subscriber PC.

PPPoA Protocols

The PPP packet includes two vital protocols as well as the user traffic—LCP and one of several types of NCPs. The LCP establishes, configures, tests, maintains, and terminates the data-link connection to ensure that the line quality is suitable for transmitting L3 protocols. A second protocol used by PPP is one of several types of NCPs. This comes into use after the data link is established by the LCP. NCP establishes and configures the applicable network-layer protocols, because PPP is designed to allow the simultaneous use of multiple network layer protocols. The NCP contains information about the higher-layer protocols, including IP and IPX. For example, the PPP NCP for IP is IPCP.

PPPoA Protocol Stack

The general PPPoA protocol stack is shown in Figure 3-10.

Starting from the user's point of view, the PC transmits an Ethernet frame that carries an IP packet. The ATU-R receives the Ethernet frame, removes the Ethernet header and trailer, and adds a PPP header. In addition to the user data, the ATU-R includes specific PPP information, such as security parameters for transport across the ADSL loop and the DSLAM.

If the CO DSL termination point is a Layer 2 multiplexer, a DSLAM, it forwards the ATM cells to the L3-enabled aggregator. If the receiving CO device is an IP DSL Switch, Layer 3 capabilities are onboard, and the following step is carried out inside the IP DSL Switch itself.

The L3 device reassembles the ATM cells into the AAL5 PDU format in the SAR process. The device then verifies and removes the AAL5 trailer, followed by verifying and removing the PPP header. Now that the data is back in IP packet form, the L3 device (IP DSL Switch or aggregator) routes the IP packet to its destination. In some cases, this might mean reconverting the IP packet to ATM cells for transmission over an ATM network, which is carried out through the same device's SAR process again.

IP Addressing in PPPoA

The user's PC and the ATU-R both require IP addresses. These addresses can be either supplied dynamically through DHCP and IPCP or applied statically.

In the PPPoA architecture, IP address allocation for the subscriber DSL modem uses IPCP negotiation, the same principle of PPP in dial mode. IP addresses are allocated depending on the type of service a subscriber uses. If the subscriber has only Internet access from the NSP, the NSP terminates those PPP sessions from the subscriber and assigns an IP address. The options for the user's PC IP address include dynamic and static addressing. IP addresses may be dynamically allocated either from a locally defined DHCP pool on the ATU-R or from the RADIUS server in the central office/exchange.

For static addressing, the service provider might provide a set of static IP addresses to the subscriber (and ATU-R) and might not assign IP addresses dynamically when the subscriber initiates the PPP session. In this scenario, the service provider uses the RADIUS function only to authenticate the user.

The PPP sessions can also be tunneled (using L2TP) from the aggregator to the final destination or home gateway router, which eventually terminates the PPP sessions. The final destination or home gateway router negotiates IPCP and can provide an IP address dynamically to the ATU-R. Static addresses are also possible in tunneling as long as the final destination has allocated those IP addresses and has a route to them.

The PC may obtain an address through DHCP from the ATU-R. This requires a DHCP address pool configured on the ATU-R. Also, the ATU-R must be configured either for DHCP relay as an intermediary between the PC and the service provider's DHCP server or as a DHCP server of its own. Either DHCP designation can be combined with NAT/PAT on the ATU-R for the subscriber PC with PPPoA.

When the ATU-R is configured for AAL5SNAP, as in bridging and PPPoE, one PVC can be used to carry multiple protocols, such as combining IP and IPX. If the AAL5MUX configuration is used, as in PPPoA, each PVC is dedicated to a single protocol, such as only IP or IPX.

Advantages and Disadvantages of PPPoA

The advantages of PPPoA are as follows:

• The architecture inherits most of the advantages of PPP used in the dial model.

• A PPPoA implementation involves configuring the ATU-R with PPP authentication information (login and password). This is a major advantage of PPPoA over simpler bridging implementations such as IRB and RBE, which have no security. That is, PPP enables per-session AAA. Per-session AAA allows the service provider to charge the subscriber based on session time for various services offered.

• IP address conservation at the DSL modem allows the service provider to assign only one IP address per subscription. When the ATU-R is configured for NAT, all users behind that ATU-R can use a single IP address to reach different destinations. IP management overhead for the NAP/NSP for each individual user is reduced while conserving IP addresses.

• NAPs/NSPs provide secure access to corporate gateways without managing end-to-end PVCs, using Layer 3 routing or L2F/L2TP tunnels. Hence, service providers can scale their business models for selling wholesale services.

• Troubleshooting individual subscribers is another advantage of PPPoA. The NSP can easily identify which subscribers are on or off based on active PPP sessions rather than troubleshooting entire groups, as is the case with bridging architecture.

• The NSP can oversubscribe by deploying idle and session timeouts using the industry-standard RADIUS function for each subscriber.

Disadvantages of PPPoA include the following:

• The relative complexity of the routing configuration on the more-capable ATU-R for PPPoA increases both labor and hardware costs.

• PPPoA has only a single session per VC. Because the username and password are configured on the ATU-R, all users behind the DSL modem for that particular VC can use only one set of services. Users cannot randomly select different sets of services, but you can use multiple VCs and establish different PPP sessions on different VCs. Of course, using multiple VCs increases configuration complexity, further increasing labor costs and the need to train service provider support personnel.

• The service provider must maintain a database of usernames and passwords for all subscribers. If tunnels or proxy services are used, the authentication can be done on the basis of the domain name, and the user authentication is done at the corporate gateway. This reduces the size of the database that the service provider has to maintain.

• If a single IP address is provided to the CPE and NAT/PAT is implemented, certain applications, such as IPTV or others that embed IP information in the payload, will not work.

• One other disadvantage of PPPoA is relatively minor for modern, highly efficient processors: Point-to-point subinterfaces require more router CPU resources (cycles).

PPPoA Implementation Considerations

In general, PPPoA is indicated by the following factors:

• Host-based software is not desired or is impossible

• Authentication and accounting are desired

• Intelligent DSL modems are to be used

• IP address conservation is important

• End-user scalability must be allowed

Labor costs, including training, deployment, and configurations of modems and central office equipment, must be factored into the costs of any PPPoA implementation.

PPPoA Summary

PPPoA (RFC 2364) uses AAL5 as the framed protocol and supports all ATM mapping options.

PPPoA is the architecture of choice for many service providers because it is highly scalable, for both the provider and the end users, and it enables L3 functionality, including security.

Considerations for deploying PPPoA include added costs of the higher functionality.