I worked on various exciting projects, such as, cross-layer optimization, inter-network cooperative optimization, femtocell architecture, interference management for wireless networks, and fault localization and scheduling in the backhaul transparent all optical network (AON). My current research interests include cross-layer adaptive transmission and optimization, resource allocation, decision making and packet scheduling techniques for wireless broadband networks, cognitive radio networks, cooperative relay networks, femtocell networks and sensor networks. He has interest in the applications of statistical signal processing, optimizations, decision and queuing theories as well. I obtained my Ph.D. under the supervision of Prof. Vijay K. Bhargava from the University of British Columbia in Aug 2007. I worked as a Research Assistant and later as a Postdoctoral Fellow with Prof. Bhargava. We investigated the cross-layer adaptation, scheduling and resource allocation issues for different wireless network over multiple input, multiple output (MIMO) channel. We used the mathematical tools of Dynamic Programming (DP), Markov Decision Process (MDP), Partially Observable Markov Decision Process (POMDP), Semi-Markov Decision Process (SMDP), Game Theory, Queueing Theory, Reinforcement Learning and Statistical Signal Processing techniques. We used Markov Desicion Process (MDP) based framework to find the optimal transmission policy. The objective was to optimize transmission power, throughput, delay and overflow. We consider the channel fading, buffer dynamics and traffic randomness. We devised an optimal policy that consider all of those information and schedule packet dynamically with fading, buffer content and packet arrival. The optimal policy gives the modulation rate, the channel coding rate and the transmission power level so that above objectives are fulfilled. We used constrained MDP formulation, where one objective is optimized keeping others below respective bounds. Constrained MDP problem is then solved using constrained linear programming techniques using interior point method. In addition to Adaptive Modulation and Coding (AMC) in the Physical (PHY) layer, our scheme consider Automatic Repeat reQuest (ARQ) along with packet combining in the Medium Access Control (MAC)/Data Link layer. We worked on the adaptation and resource allocation issues in the Wireless Sensor Network and the OFDM-based Cognitive Radio Network as well. Along with optimal policy, we also devised easier to implement suboptimal and heuristic based policies, and compared their performance with the optimal policy. Whereas optimal policy acts as a benchmark, suboptimal policy is easier to implement and computationally less extensive. We used benchmark to judge the performance of suboptimal policy. We found that our devised policies performing very well and that their performance very close to the optimal policy in finding policy in a certain situation. In our work, we also consider the cases when the channel state and/or traffic states may not be fully observable. Hybrid ARQ techniques have been proposed in different wireless standards, including IEEE 802.11 (WiFi), IEEE 802.16 (WiMAX), 3GPP and 3GPP2. In our work, we consider hybrid ARQ using both SW-ARQ (Stop-and-wait ARQ) and SR-ARQ (Selective-repeat ARQ) techniques. Please visit publication page to view my works. |