Simplifying the configuration of 802.11 wireless networks with effective snr

摘要: Advances in the price, performance, and power consumption of Wi-Fi (IEEE 802.11) technology have led to adoption wireless functionality diverse consumer electronics. These trends enabled an exciting vision rich applications that combine unique features different devices for a better user experience. To meet needs these applications, network must be configured well provide good performance at physical layer. But because usage trends, finding configurations is increasingly challenging problem. Wireless configuration objectives range from simply choosing fastest way encode data on single link global optimization many interacting parameters over multiple sets communicating devices. As more links are involved, as advances (e.g., OFDM MIMO techniques Wi-Fi), used changing channels, size space grows. Thus algorithms find operating points among growing number options. The heart every algorithm evaluating particular point. For example, if we know all three between source, destination, potential relay, can easily determine whether or not using relay will improve aggregate throughput. Unfortunately, two standard approaches this task fall short. One approach uses signal strength statistics estimate but do yield accurate predictions performance. Instead, practice measures by actually trying possible configurations. This procedure takes long time converge hence ill-suited large spaces, devices, which today. result, complexity practical dominated optimizing estimation step. In thesis, I develop comprehensive rapidly accurately predict point space. devise simple powerful model low-level channel measurement extrapolates wide My work makes most complex step today's algorithms—estimating effectiveness configuration—trivial, achieving enabling solution larger problems.