Online Channel Measurement And Application In Mobile Wireless Network

Themobilewirelessnetworkshaveexperiencedrapiddevelopmentinrecentyears,and the growth is expected to continue unabated. For diferent types of mobile net-works, a basic consideration is the accurate channel state estimation and link qualitymeasurement to get efcient trade-of between reliability and data rate. The most im-portant factors for mobile wireless networks are channel state and link quality, both ofwhich play signifcant role in real-time operating. Therefore, accurate channel mea-surement and prediction is becoming increasingly important in mobile wireless net-works. It is crucial to reduce the estimation overhead to address the issues of mobilityand multi-confguration. As for mobile wireless networks, another issue for onlinemeasurement is to be dynamic to network status resulting in suitable trade-of betweenreliability and data rate.For GSM-R networks deployed for high-speed railway, the sampling frequencyis high in current methods which leads to high cost. These methods can only be usedin drive test but not for online monitoring. Here comes the problem that how to makeaccurate channel measurement with lower overhead. But the radio propagation envi-ronments and the varied terrains along the high-speed railway route are complex. Theonline measurement algorithm should be adaptive to these changes and reduce over-head with guaranteed accuracy.Ontheotherhand, thecontinuedsuccessofmobile802.11ndependsontheirabil-ity to efciently confgure diferent PHY/MAC enhancements. This is challenging inthat multi-confguration in mobile802.11n not only requires far more samples to ac-quire sufcient information for all possible channel settings, but also introduces signif-icant complications in channel modeling. Furthermore, channels are more vulnerable to environmental variability and terminal mobility in mobile802.11n. This is furthercomplicated by the distinctive features in mobile802.11n networks, specifcally, mul-tiple PHY/MAC settings and spatial-temporal variation channels. Therefore, the accu-rate prediction of link quality should be adaptive to mobility and multi-confguration,so that the network performance can be improved.This paper proposes the online measurement algorithm for GSM-R and802.11nnetworks respectively. To deal with the issues of mobility and multi-confguration, theonline channel estimation and link quality modeling are presented, which reduce themeasurement overhead with guaranteed accuracy. Through the real world implemen-tation on our testbed, the proposed measurement methods and rate adaption algorithmare evaluated. The experimental results show that the online measurement algorithmsignifcantlyreducesthemeasurementoverheadwithguaranteedaccuracy, andthenet-work performance including packet delivery and throughput are also improved.