Evaluating Packet-level Forward Error Correction: 1-D interleaved parity codes

Packet-Level Forward Error Correction (PL-FEC) techniques are widely applied to multimedia communications. To deal with the burst loss problem in packet transmissions over networks, PL-FEC 1-D interleaved parity codes are used in Real-time Transport Protocol (RTP), which uses User Datagram Protocol (UDP) to transport packets. However, the codes´ error correction performance, encoding rate, decoding rate and throughput have not been comprehensively evaluated recently. Besides, the width of the interleaving window is not specified in corresponding RTP; unsuitable window width may result in damaging the codes´ error correction performance. In this paper, the encoding and decoding algorithms were given, and the algorithms were implemented by programming 1-D interleaved parity codes´ encoder and decoder on computers, which produced high encoding and decoding rates. Gilbert and Elliott (GE) model was used for simulating the losing packets behavior. In real experiments, UDP was used for the transport of a file of H.264 encoded by 1-D interleaved parity codes´ encoder. By simulative and real experiments, the codes were evaluated; the suitable window width was suggested. The codes coding are simpler than other PL-FEC codes coding. Besides, a simple comparison of the codes and Reed-Solomon (RS) codes was done; for multimedia communications, the codes are better than RS codes.