RELIABILITY EVALUATION OF A PACKET-LEVEL FEC BASED ON A CONVOLUTIONAL CODE CONSIDERING GENERATOR MATRIX DENSITY

Abstract

AbstractAs the Internet and its many applications become pervasive throughout the world, packet loss recovery is becoming an important technique for the reliable transmission of data. There are two types of packet loss recovery: the automatic repeat request (ARQ) and the forward error correction (FEC). In ARQ, receivers send acknowledgement messages and request packet retransmissions from senders. On the other hand, FEC employs proactive redundant packets without any retransmission. From the viewpoint of real-time applications, FEC is considered an extremely promising technique since there are no time-delays when retransmitting data. Several researchers conducted a study of Read-Solomon-code-based packet-level FEC's. We have shown that a convolutional-code-based packet-level FEC is more efficient under the low packet loss ratio. In this paper, we examine convolutional-code-based packet-level FEC's considering the density of the generator matrix. Stochastic analysis and simulations show the effect of our new FEC scheme.