In order to meet very low packet loss rate (PLR) requirements under strict delay constraints for IP based DVB services over wireless home networks (WHN), an optimized real-time transport protocol (RTP) level hybrid error correction (HEC) scheme is proposed in this paper. The scheme is compared with the Adaptive Forward Error Correction (AFEC) scheme using Reed-Solomon code. By the HEC scheme, the needed redundancy information (RI) for repairing missing packets can be minimized, especially in DVB systems with a single user. The result of quantitative analysis shows that the needed RI of the proposed scheme can even achieve the Shannon limit with only three retransmission rounds. On the other hand, our studies show that the performance of the HEC scheme is better than that of the AFEC scheme if both the average link PLR and group size are small. However, if both the average link PLR and group size are large enough, the AFEC scheme performs better than the HEC scheme. At last, the simulation results for one typical scenario are also provided in the paper to validate the result of quantitative analysis for the HEC scheme. As a result, because of the simplicity and efficiency of the proposed HEC scheme, it is a very good choice for guaranteeing the Quality of Service for DVB systems in WHN with the small group size and small average link PLR. However, for scenarios with the large average link PLR and large group size, the AFEC scheme should be taken into account.Index Terms-ARQ, DVB, forward error correction, ReedSolomon code, wireless home networks.

No abstract

In this paper, an application layer Hybrid Error Correction scheme with Reed-Solomon (HEC-RS) code is proposed to satisfy the very low Packet Loss Ratio (PLR) requirement under a strict latency limit for DVB services over wireless LANs. The PLR performance of the HEC-RS scheme is analyzed and compared with the HEC scheme based on the Packet Repetition (HEC-PR) technique and an Adaptive Forward Error Correction (AFEC) scheme based on RS codes. It is found that the needed Redundancy Information (RI) of the HEC-RSscheme is much less than that of the AFEC scheme. However, if the group size and link PLR are small enough, the needed RI of the HEC-RS scheme is proven to be much more than that of the HEC-PR scheme. As a result, the optimum performance can be achieved when combining the HEC-PR and the HEC-RS scheme.

No abstract

No abstract