Outage Performance Of Diversity Over Composite Fading Channle With Shadowing Correlation

The performance evaluations of wireless communications systems are significantly affected by the selection of channel models, and the precise and effective of evaluations are always accompanied with the accurate and complete channel models. There is part of a more general trend suggesting that most channel models used in simulation and analysis work today are sometimes too simplistic and may lead to misleading results. To overcome such flaw, it is necessary to select channel models which are more accurate and complete, and well-agreed with the practice. On one hand, in the scenarios with both macro-diversity and micro-diversity, such as the Distributed Antenna Systems, the composite shadowed fading channels draw a great deal of interests now. On the other hand, compared with the correlation among multipath fading, the correlation among shadowing are the dominant sffects on the performance of composite diversity. Therefore, in this paper, we focus on the outage performance evaluation issues of OSC (Optimal Selection Combining) and MRC (Maximal Ratio Combining) diversity over such composite shadowed fading channels with shadowing correlation.Firstly, the methods for the outage performance evaluations of diversity over pure correlated lognormal channels are reviewed as a foundation. Secondly, the outage perfor-mance evaluations of diversity over composite shadowed fading channels are handled, and the corresponding solution approaches are proposed. In particular, on the condition that the multipath fading severity measure m in the Nakagami-m fading is an integer, a simple method is presented to approximate composited shadowed fading channels with shadow-ing correlation by pure correlated lognormal channels, and consequently, the desired eval-uation issues over composite channels can be translated to the issues over pure lognormal channels; on the condition that m is an arbitrary positive value, two numerical methods ire propose to solve the desired evaluation issues, and meanwhile, the accuracy and effi-ciency of these two methods are analyzed, as well as the selection guideline for these two methods in different applications is presented. Last but not the least, through utilizing the above-proposed approaches, the outage performance evaluation issues in hybrid diversity systems over composited shadowed fading channels are handled, and effective solutions are introduced. Through the demonstrated simulations in typical distributed antenna sys-tems, the implied effects of corresponding parameters on the desired outage performance are revealed and analyzed. The main contributions of this paper are as follows.In order to solve the issues of outage performance evaluations over composite chan-nels with shadowing correlation on the condition that the multipath fading severity mea-sure m in the Nakagami-m fading component is an integer, we proposed a simple manip-ulating method to approximate such composite channels with shadowing correlation by pure correlated lognormal channels, and through this method, the desired evaluation is-sues over composite channels can be translated to the issues over pure lognormal channels which have been already well solved. In particular, based on the method of approximate a single Nakagami-m/Lognormal (or Rayleigh/Lognormal) random variable (RV) by new lognormal RV and through the joint matching approach, the multivariate composite Nakagami-m/Lognormal (or Rayleigh/Lognormal) distribution are approximated by a new multivariate lognormal distribution with the relationships between original and approxi-mated parameters solved and the accuracy of such approximation validated and analyzed by simulations. It is worth to note that, besides using such approximating method to solve the desired performance evaluation issues, this method can also be utilized in other issues aver composite channels with shadowing correlation. Due to the simplicity of the compos-te channels and probability of directly utilizing the existing results over pure lognormal channels, the potentials of this approximation in wireless applications can be predicted.In order to solve the issues of outage performance evaluations over composite chan-nels with shadowing correlation on the condition that the multipath fading severity mea-sure m in the Nakagami-m fading component is not an integer, the method of approx-imating the sum of correlated lognormal RVs by a single lognormal RV is extended to the method of approximating the sum of correlated composite distributed RVs by a sin-lie lognormal RV, and then the accurate and effective evaluation solution is proposed. In particular, through the moment generation function (MGF) based matching, and relation-ship between the sum of original RVs and the approximated RV is well solved, and then the desired performance evaluation issue is readily translated to be the issue over single lognormal channel. It is worth to note that, besides using such approximating method to solve the desired performance evaluation issues, this method can also be utilized in other issues with regards to the sum of correlated composited distributed RVs.For the numerical evaluations for the outage performances of diversity over com-posite shadowed fading channels with shadowing correlation, the Genz algorithm is in-troduced to enhance the calculation efficiency on condition of larger diversity order (or squivalent, diversity branches number). The desired performance evaluations display a multi-fold integral form due to the multivariate lognormal distribution, and the traditional ipproach is to numerical calculate it with the Gaussian-Hermite polynomials. However, when the diversity order increase, the computation burden will increase exponentially and the calculation efficiency decrease significantly. To overcome this flaw, a new method called Genz algorithm is introduced, and this algorithm has the ability that the computa-tion burden will increase linearly with the diversity order increasing. Through extensive simulations, the accuracy and efficiency of those above-mentioned methods are validated and analyzed, respectively. More importantly, results prove the Genz algorithm, which :an not only satisfy the higher accuracy requirement but also maintain lower computa-tion burden, is better than the Gaussian-Hermite polynomials on the condition of larger diversity order. It is concluded that, the introduction of Genz algorithm can solve the de-sire performance evaluation issues effectively, especially for the scenarios in the Massive Distributed Antenna Systems.For the outage performance evaluation of hybrid diversity over composite shadowed fading channels with shadowing correlation, we propose corresponding solutions for dif-ferent hybrid diversity strategies. Specially, for the hybrid diversity with both OSC macro-diversity and MRC micro-diversity, the solution to the desired performance evaluation is proposed with both correlation among shadowing and correlation among multipath fading taken into account. On one hand, for the " AA" type of hybrid diversity (both the macro-diversity and micro-diversity using the same combining strategy), through the correlation matrix adjusting, the desired performance evaluation issues can be naturally solved by the general issues over composite shadowed fading channels with shadowing correlation. On the other hand, for the "AB" type of hybrid diversity, through considering the out-put of micro-diversity as the input of macro-diversity, the desired performance issues can be equivalently solved by the general issues over composite shadowed fading channels with shadowing correlation. Simulation results validate the accuracy and effectiveness of these solutions. In the meanwhile, a typical Distributed Antennas System is demonstrated is an instance to analyze and reveal the implied effects of corresponding parameters on the desired performance.