New algorithms to generate balanced workload clusters with reduced communication

Load balancing is a technique to distribute workload evenly across two or more computers,network links,cpus,hard drives,or other resources,in order to get optimal resource utilization,maximize throughput,minimize response time,and ^void overload. when clustering objects to be allocated on a number of processing units,most research focus on either balancing of the workload among the clusters or minimising the communication among them. in this paper,we developed three clustering algorithms that combine both objectives: balancing the load of different clusters of objects while minimizing the communication among them. we show that communication graphs of these hybrid algorithms h-k partition algorithm and h-k medoids algorithm methods have a good partition that can be found in distributed technique particularly in the major clusters and objects number. in the other hand,the d-k partition algorithm and d-k medoids algorithm occupy the second level.