The Resource Management And Spreading Model Of The Mobile Learning System Over CDMA

With the progress of society, lifelong study is becoming more and more necessary for citizens: no matter it is for obtaining a degree, or renewing their knowledge and improving their technical ability. Sometimes, the study is not only an effort to get promotion in their career, but also an effort to maintain their job. Distance learning is the best form providing lifelong study and mobile learning because it provide learning with more freedom. As is known, mobile communication and Internet are the two among the fastest growing areas in the contemporary telecom worlds. Their combination results in the emergence of mobile data communications and the mobile Internet. The fact that electronic commerce has become mobile commerce makes us know that mobile learning is coming to us nearer.In this paper, the carrier format of mobile Internet, based on the combination of mobile cellular infrastructures and Internet, is employed to construct a network model architecture of mobile learning system over CDMA, then the paper discusses the construction of software and hardware, and key technology, and propose constructing a model of learning resource organization and propagation model, finally constructs an experimental system model.It adopts the criterion of IEEE Learning Technology Standardization Committee's Learning Objects Metadata (IEEE LTSC LOM) in the model of learning resource organization. As mobile learning is a typical mobile computation, it adopts Tree-Tier Replication Architecture in resource base model, then designs the air replication model and mobile client caching model which is suitable for the characteristic of mobile learning.Owing to the first tier replication-replication of server-it enables many servers available to every client, whereas the stat bus of the available severs is dynamic. However, the bandwidth of network infrastructure available now is limited, so it is a key problem that how to select the replica server dynamically and provides the best service to clients at their request.This paper proposes a Client-Centered Load Distribution mechanism which is different from the conventional algorithm of selecting dynamic replica servers. Rather than binding a client to its "most convenient" replica server, the mechanism intercepts the request to visit a Web page of each client browser into fragments. Each sub-request is issued to a different available replica server, concurrently. The replies received from the replica servers are reassembled at the client end to reconstruct the requested page, and then delivered to the client browser. It's preferable to realize C2LD mechanism on the WAP gateway.