ENERGY AWARE REINFORCEMENT LEARNING NETWORK SECURITY
Journal: JOURNAL OF COMPUTER ENGINEERING & TECHNOLOGY (JCET) (Vol.8, No. 1)Publication Date: 2017-02-26
Authors : S. SENTHIL KUMAR; T. N. RAVI;
Page : 60-72
Keywords : Artificial Intelligence; Reinforcement Learning; Markov’s Decision Process; StateAction-Reward-State-Action; Energy Aware Reinforcement Learning Network Security;
Abstract
Reinforcement Learning (RL) is an Artificial Intelligence (AI) procedure which performs better than other AI procedures in case of Network Security. Adding some Energy Awareness to this RL method is proposed in this work. Many of the participating nodes of modern computer networks are battery operated. Markov's Decision Process (MDP) is used to provide energy awareness to Reinforcement Learning method and this automated security system can switch between a number of ‘various resource consuming security protocols' to maintain a delicate balance between security establishment and power conservation. The power and security balance is further maintained using State-Action-Reward-State-Action Algorithm (SARSA). The power and security balancing scheme is named as Dynamic Power Management (DPM). This Energy Aware Reinforcement Learning Network Security (EARLNS) handles multiple communication data categories ranges from minimum bytes control data transfers to maximum bandwidth starving multimedia data without compromising Security
Other Latest Articles
- EXTENDING BAYESIAN LEARNING BASED NEGOTIATION TECHNIQUE TO A TWO-ISSUE BILATERAL NEGOTIATION IN CLOUD
- MEDICAL DIAGNOSIS CAD SYSTEM USING LATEST TECHNOLOGIES, SENSORS AND CLOUD COMPUTING
- FRAMEWORK FOR FEEDBACK ASSESSMENT AND ASSISTANT E-SYSTEM IN DIGITAL AGE EDUCATION
- SERVICE CHECK: EFFECTIVE USER BEHAVIOR ANALYSIS & TRUST MANAGEMENT – IDENTIFY BEST CLOUD
- A COMPARISON OF ESTIMATED LIFE OF KNEE IMPLANT APPLYING THE MULTIAXIAL FATIGUE CRITERIA
Last modified: 2018-09-20 15:29:01