Prediction based Dynamic Load Balancing System for Text to Speech Conversion

Dynamic load balancing becomes a prominent stakeholder in a distributed system for execution of massive computational load. Based on the dynamic state of underlying resources dynamic load balancing disseminates load efficiently to underlying networked resources to achieve efficient and unfair load balancing. It's very difficult task to choose scheduling mechanism in these cases. To predict the load of incoming jobs accurately and matchmaking of these jobs to computational resources dynamically is very crucial task. In proposed research we have used linear regression mechanism for achieving dynamic load balancing in text to speech conversion process. Now a day's shrewd users prefers listening the text literature than reading it. The considerable amount of load is generated on the system as number of input text files increases. Proposed capacitive dynamic load balancing technique considers the computational capacity of CPU which is taken as a key factor for load dissemination. In this technique load given to a CPU is proportionate to its computational capacity.