Downloads & Free Reading Options - Results

For many communication applications with time constraints (e.g., transmission of packetized voice messages), a critical performance measure is the percentage of messages which are transmitted within a given amount of time after their generation at the transmitting station. In this talk we present a random access algorithm (RAA) suitable for time-constrained applications. Performance analysis demonstrates that significant message delay improvement is attained, at the expense of minimal traffic loss. We also consider the case of noisy channels. The noise effect appears at erroneously observed channel feedback. Error sensitivity analysis shows that the proposed random access algorithm is insensitive to feedback channel errors. Window Random Access Algorithms (RAAs) are considered next. These algorithms constitute an important subclass of Multiple Access Algorithms (MAAs); they are distributive and they attain high throughput and low delays by controlling the number of simultaneously transmitting users. Theses. (FR)