Lossless Parallel Adaptive Compression for Satellite Networks

Traffic on today's communications networks of interconnected Local and Wide Area Networks is limited mainly by the bandwidth of the communication channel. Increasing bandwidth to accommodate the high data rate of multi-tactic transmissions is necessary. One way to balance the situation is to harness available local processing power to compress data before it is transmitted. We propose an alternative system in which the compression process is offloaded to a dedicated processor to allow real time compression while minimizing the impact on local system performance. Digital signal processing techniques are used to implement real time adaptive compression algorithms to reduce the volume of data to be transmitted. Multiple algorithms for entropy encoding are run in parallel to produce results from which the best compression of a stream of data may be selected. The networking of computers via satellite systems introduces a particular subset of problems that must be considered for real time compression of transmitted data. To produce a robust lossless compression system, the interaction of various elements of the communication system such that transmission protocols, protocol headers, data encryption, and timing delays, are investigated.