With the rapid development of broadband wireless networks, more and more attention has turned to distributing real-time multimedia services over wireless networks. Recently, Wireless LANs (WLANs) based on IEEE 802.11 have come into widespread use for accessing the Internet and are expected to be used for Digital Video Broadcasting (DVB) services over home and nomadic networks. On the other hand, since IP multicast provides a scalable and efficient means for distributing datagrams to a group of receivers, IP-based networks were proposed for delivering DVB services. The DVB systems based on IP multicast typically employ an application-level protocol to provide some information about the set of receivers and reception quality statistics. The Real-time Transport Protocol (RTP) is usually used for this purpose. RTP or MAC layer of the IEEE 802.11 does not, however, guarantee any Quality of Service (QoS) for real-time multimedia multicast applications, although packets lost usually happen due to collisions, fading or interference in WLANs. Therefore, it is essential to employ some error control techniques at application layer to guarantee very low Packet Loss Ratio (PLR) requirement needed for DVB services over wireless home networks.

Unlike non-real-time services, real-time DVB service is very sensitive to the end-to-end delay of the audio/video packets. That is, the real-time audio/video packets have deadlines by which they must be delivered to the destination; otherwise, those frames have to be discarded resulting in the distortion of the audio/video. Therefore, the error control techniques employed for DVB services must consider strict delay constraints.

On the other hand, as well known, there are mainly two categories of error control techniques at the application layer:

• Automatic Repeat Request (ARQ) that retransmits the lost packets
• Forward Error Correction (FEC) that transmits redundant parity packets

Many studies have shown that Hybrid Error Correction (HEC) method that combing ARQ and FEC can perform better than ARQ or FEC alone. We thus focus on the performances of HEC schemes for DVB services in wireless home networks. Actually, the project includes three parts as follows:

1. Develop a mathematical framework to analyze and optimize HEC schemes under strict delay constraints
2. Validate the theoretical results by simulation using Network Simulation Tool: ns2
3. Implement a practical system (DVB services over IEEE 802.11/a/e) for validating the theoretical results

In the following, we will take three simple examples for explaining the principle of FEC, ARQ and HEC techniques. First, Figure 1 shows an example of operation of an FEC system. In this example, the FEC encoder produces two redundant parity packets (P1, P2) for every three data packets sent. If one data packet (D3) and one parity packet (P2) are lost, the receiver can reconstruct the data packet (D3) by FEC decoder.

Finally, Figure 2 gives an example where different packets are lost for different receivers in the first transmission for a multicast scenario. As shown in figure, two different mechanisms can be used for recovering the lost packet in different receivers. In scheme ‘A’, error recovery by multicast retransmissions of the original data packets requires retransmission of all lost packets, which acts as a pure ARQ scheme. In scheme ‘B’, retransmission of a single parity packet allows all receivers to recover their lost packet, which acts as a HEC scheme.

• Manuel Gorius