mINCARNATE: An interference and mobility aware spatial scheme for tightly coupled LTE–Wi-Fi networks

Abstract

Stochastic geometry has clinched notability in the past several years. It is a robust mathematical tool for analyzing wireless systems due to its tractability in nature. In this work, a flexible and docile model has been proposed for heterogeneous wireless networks consisting of tightly coupled Long Term Evolution (LTE) Small cell eNodeBs (SeNBs) and Wireless Fidelity (Wi-Fi) Access Points (APs). By the assistance of stochastic geometry, the key performance metrics of LTE–Wi-Fi Aggregation (LWA) system have been analyzed in non co-located scenario. The positions of SeNBs and APs are modeled as two independent homogeneous Poisson Point Processes (PPPs). Enabling LWA operation with an arbitrary number of Wi-Fi APs in a given region may not ensure maximum rate and coverage for the mobile operators. A novel scheme, coined as, InterfereNCe Aware matéRN hArd-core poinT procEss (INCARNATE) has been proposed to increase the performance of LWA system by allowing LWA operation with a chosen subset of Wi-Fi APs available at the disposal. We derive Signal-to-Interference-plus-Noise Ratio (SINR) distribution of UEs which are associated with SeNB, AP or LWA node (i.e., SeNB and AP). This further helps to find out the joint coverage probability and average data rate over the network. In addition to the density of APs, the velocity of UEs plays a vital role in analyzing Key Performance Indicators (KPIs) of the system. So, INCARNATE scheme has been further extended to mINCARNATE scheme wherein mobility of UEs is introduced in the LWA system. A handover model of LWA system has been proposed by considering the mobility of the UEs. Expected number of handovers and average data rate have been analytically measured. Average number of handovers observed per UE and throughput of UEs have been empirically evaluated by varying velocity of UE. Further, cost is defined as a function of velocity of UEs, number of handovers, and density of deployment. Capital Expenditures (CAPEX) and Operational Expenditures (OPEX) can be minimized for a given user velocity by operating at minimal cost value. INCARNATE scheme outperforms the traditional MHCPP scheme by 73% and 17% in terms of data rate and coverage probability, respectively. Similarly, LWA with INCARNATE scheme excels by 51% and 6.23% as compared to regular LWA in terms of data rate and coverage probability, respectively. The proposed cost function assists to obtain the optimal deployment of LWA nodes using mINCARNATE scheme. LWA with mINCARNATE scheme improves the throughput by 53% as compared to native LWA handover scheme.