Exploring modelling, analysis and resource allocation in computer and communication networks.

Exploring modelling, analysis and resource allocation in computer and communication networks.

Used performance analysis and control tools to design scheduling algorithms for transmission of streaming video traffic over wireless channels. The objective was to minimize video quality switches while being fair in other performance measures such as throughput.

Investigated the requirement of performance management in IEEE 802.11 infrastructure WLANs. Proposed a coarse time-sliced approach to achieve fairness in WLANs. Proposed heuristics for link dependencies, and an on-line rate adaptation algorithm leveraging the closed loop dynamics of TCP.

Investigated the requirement, and the problem of aggregate utility maximization in wireless mesh networks under a distributed greedy heuristic. Provided insights into the factors affecting aggregate utility maximization in a network, by providing bounds on the same. Quantified the sub-optimality of the greedy schedule by looking at the convergence properties of the \(\epsilon\)-subgradient method.

Studied the problem of joint congestion control, routing and MAC layer scheduling in multihop wireless network in the framework of cross-layer optimization. Explored the use of a known centralized greedy heuristic, and developed a distributed algorithm that can schedule independent links based on local information. Studied convergence issues related to the price update algorithm. Developed and simulated a practical protocol that maximizes aggregate utility in a wireless mesh network. Also studied delay incurred by application packets sent over a WLAN.

Helping streamline process and policies with the objective of meeting the scientific and technological mandates of the company.

Focused on scientific and technical issues within the company. Essentially, was responsible for leveraging the right technology, scientific knowledge and system architecture to create a market ready product.

Built a Java based discrete event simulator to evaluate the performance of the carrier allocation methods. Proposed and implemented a probabilistic demand forecast module. Proposed and implemented a carrier predication and segregation module. Proposed and built a website to demonstrate the efficacy of the above mentioned modules.

As a part of this startup, my role was to investigate, identify and propose methods to tackle on-demand vehicle routing problem. The proposed methods were also incorporated into a web portal using technologies such as Java, Javascript, Node.js and MongoDB.