Power Aware Software(current research area)
Thesis Title: Power Aware Software
Supervisor: Professor
Ajit Pal
, Department of
Computer Science and Engineering
,
Indian Institute of Technology, Kharagpur
Abstract:
Power/Energy saving is a vital issue not only for embedded and portable devices but also for servers, personal, mainframe and supercomputers. It has been observed that more power reduction can be achieved at algorithmic level than at circuit and gate level. In past decades of CMOS technology dynamic power dissipation dominated the leakage power. As CMOS technology is reducing in dimensions, the leakage power is gradually becoming a challenge for power awareness. In addition to low power circuits, it is necessary for the system softwares like compilers and operating systems to be well equipped to achieve low power. Compilers should deal with code optimization techniques to reduce power consumption of a program. Operating Systems should care process, memory, I/O and file management to achieve low power. Achieving low power is challenging for real time systems as it may degrade performance. So study of power-performance tradeoff for a given program is very important. Dynamic Voltage Frequency Scaling is a handy technique for power management. Processors running on multiple voltage-frequecy pairs helps in achieving low power computation. Multicore processors have evolved keeping power and thermal management in mind. These days every computer is connected to the Internet. The network protocols in each layer of the OSI or TCP/IP stack should be enhanced to achieve low power. Network Security is an important area for power awarness, so there is a need to implement low power cryptographic algorithms.
Network Mobility(research area for Master's dissertation)
Thesis Title: Pinball Routing Elimination with Route Optimization (PRERO) in Multihomed Mobile Network
Supervisor: Professor
Sulata Mitra
, Department of
Computer Science and Technology
,
Bengal Engineering and Science University, Shibpur
Abstract:
A solution of the pinball routing problem in a nested mobile network is proposed in the present work. All the communication to and from the mobile network node in a nested mobile network passes through the tunnel between mobile router and its home agent as mentioned in network mobility route optimization problem statement. Such problem is formally known as pinball routing problem. But in the present scheme the communication to and from the mobile network node may or may not pass through the home agents of mobile routers at the higher level of the nested network depending upon the network mobility. Such consideration helps to minimize delay due to pinball routing effect. Moreover in the present work direct communication is proposed if both the source node and the destination node lie inside the same nested mobile network. The proposed scheme is the combination of route optimization algorithm and pinball routing elimination algorithm in a nested multihomed mobile network. The route optimization algorithm selects an optimal route inside the mobile network. The pinball routing elimination algorithm selects the pinball eliminated route from the mobile network to the correspondent node. The performance of the proposed scheme is evaluated on the basis of tunneling overhead, end to end packet delay and fragmentation using NEMO_SIM simulator. Results based on a detailed performance evaluation study are also presented to demonstrate the efficacy of the proposed scheme.
Publications