Home Engineering Modeling and Optimization for Mobile Social Networks
Cloud Resource Allocation in MSNs
MSNs allow more users to have interactions with each other and obtain various multimedia content [1-3]. Recent studies  show that the number of users keeps increasing and the traffic of mobile data will be nearly tenfold in 2019, compared with that in 2014. Especially, with the popularity of shared data plan in the near future, users may not only obtain but also share more multimedia contents with others who have social relations with them [5-7]. Therefore, providing users with efficient multimedia services becomes more important than before [5, 8, 9].
However, to provide users with satisfied multimedia services, there exist some new problems to be resolved. On one hand, due to the explosive growth of volume of multimedia and the high demand of quality of experience (QoE), providing users with multimedia services [10-12] needs a large amount of resource. However, the local mobile devices in users always have the limited resource, such as capacity, bandwidth,
© Springer International Publishing AG 2016 75
Z. Su et al., Modeling and Optimization for Mobile Social Networks,
buffer, etc. New consideration is needed to reduce the consumed resource. On the other hand, multimedia content servers are remotely placed from users. It takes time for users to obtain the requested multimedia content, resulting in a further QoE degradation. For example, if a user wants to watch a movie with his mobile device, the content of movie has to be retrieved from a remote multimedia content server through a large number of routing nodes.
To resolve the above issues, media cloud has been advocated with the following reasons . Firstly, media cloud can deploy cloud resource to process multimedia tasks. Some complicated computations or large-sized multimedia content storage requiring extra resource can be performed at the side of media cloud, where the required resource can be reduced for users. Therefore, the media cloud can help users to save their resource. Secondly, a broker  can be placed between media cloud and users. As the broker can act as a proxy which is close to users, users can connect media cloud through the broker for obtaining multimedia services. With the high-speed communication links between media cloud and the broker, users can obtain multimedia services faster than contacting the remote multimedia content servers.
As the resource to be allocated among media cloud, brokers and users is limited, resource allocation becomes a very important challenge. However, the conventional resource allocation schemes can not be directly used to allocate resource among these three parties. First, there exist some significant social features in media cloud with users. For example. users within the same community may have the same interest and social activities [15-17], resulting in the similar demand of QoE for multimedia services. Therefore, social features should be considered to determine the resource allocation. Besides, users in the same community can know the information of each other. Thus, the decision of a user on the selection of broker may be influenced by others. As a result, the affection and competition among different parties should also be taken into consideration for resource allocation.
Although some related studies have been carried out to study resource allocation about cloud computing and mobile networks [18, 19], few of them have studied the resource allocation problem based on the social features in media cloud. In addition, most of them mainly focus on behaviors of servers, instead of three parties including media cloud, brokers and users. Therefore, it is still a new and open problem to design resource allocation scheme of media cloud with users.
|< Prev||CONTENTS||Next >|