Development and analysis of error control techniques of wireless ATM networks

Recently wireless services have become increasingly popular. There is considerable interest in providing ATM-like virtual connectivity to the wireless domain, i.e., wireless ATM (WATM). However, the potentially highly noisy wireless environment has been questioned for the feasibility of a WATM network. Therefore, error control becomes a very important issue in the successful design of WATM networks. If the error control mechanisms can satisfy the quality of service (QoS) requirements of WATM, a WATM network will provide value-added integrated wireless services and extend a seamless access to ATM backbone networks.; Due to various conflicting requirements lying in WATM networks, finding the desired error control solutions becomes very difficult and challenging. The research here concentrates on the development and analysis of cell-level forward error correction in WATM networks.; A cooperative error control scheme for multi-class traffic protection under residual link error and node buffer overflow situations is investigated. In this scheme, unequal error protection enhances overall traffic robustness. Intelligent buffer management with a loss shaping strategy highly concentrates overflow losses into fewer data blocks to make FEC effective and retransmission efficient. Embedded error detection reduces redundancy. Simulation results demonstrate that voice, video, and data performance has been effectively improved with respect to cell loss, symbol error, and block error-loss rates.; We also propose an adaptive decoding procedure based on generalized array codes, which construct the fundamental coding frames for protecting multiple traffic classes in a wide range of applications and networks. The mathematical analyses and computer simulations show that the coding capabilities of array codes have been further significantly exploited.; To obtain the relationship among the decoding accuracy, speed, and code parameters, making a judicious code and decoding procedure selection available, software array codecs with adaptive and other decoding procedures are implemented. Performance comparisons are performed. Our running tests show that the adaptive decoding procedure is very promising in various applications.; A closed-form expression of erasure recovery by Reed-Solomon codes is also derived. It is very suitable in reducing the complexity and decoding time in the design and implementation of various types of codecs.