- Big Data Processing - Performance Gain Through in-memory Computation: The intention is to analyze and report on how the MapReduce and in-memory Spark compute engines of Apache Hadoop impact big data processing. Appropriate benchmarks are being considered to analyze their performance.
- Data mining: To classify DNA segments into two categories, supervised learning (Logistic Regression, Decision Tree etc) was used. A consensus-based and a PWM-based model were built in this project.
- Tool(s): Java, Weka, MATLAB, Shell Script
- Compiler: A complete compiler for TL13 programming language that includes a recursive descent parser, a core that outputs MIPS assembly code (intermediate output includes a type-annotated-AST, and a control flow graph labeled with ILOC instructions) with optimizations (translation into SSA form) and language extensions.
- Middleware Framework Construction: In this project a middleware to support remote objects for the Object Request Broker (ORB) architecture was implemented.
- Mobility Induced Congestion in Mobile Ad Hoc Wireless Networks: To improve the performance of TCP in mobility induced congestion scenario for MANETs, AODV routing protocol was modified. The idea of deflection routing protocol for wired network was adopted in wireless networks in two different ways.