Wireless networks possess several distinct characteristics that render the analysis, design and actual implementation methodologies developed for wired networks either invalid or inefficient. Specifically, wireless channels are error-prone, wireless links are volatile and link quality in terms of error probability and available bandwidth is time-varying. Several methods have been proposed to cope with the above issues in a wireless environment. Two well known are the Dynamic Rate Adaptation and the Forward Error Correction (FEC). The first one adjusts the transmission rate in a per packet basis based on the channel quality. The second generates and transmits redundant parity packets in order to make the communication more robust. The receiver can use block erasure codes in order to correct errors using redundant information in the data stream, without contacting the sender.
In this project we plan to combine the mentioned technologies in a joined scheme that will dynamically adjust the transmission rate and the FEC in order to optimize the decoding procedure at the receiver. Conventional systems use a conservative approach of transmission rate adaptation in order to cope with the errors in the channel. In this scheme, after a small number of consecutive packet losses the transmission rate is reduced to the lower level. The proposed mechanism uses a more aggressive approach: It keeps the transmission rate high, even for a considerably high packet error rate. However, the scheme applies the appropriate FEC in order to correct the increased errors. The joined transmission rate/FEC mechanism is efficient since by busting the transmission rates it leaves room for extra FEC and therefore it achieves higher throughput. Additionally it is a very practical mechanism since it is easy to implement. We are planning to implement the scheme in software and run experiments in a real environment.