loud computing has become one of the hottest topics in both academia and industry. Cloud computing is a model for enabling on-demand access to a shared pool of configurable resources (e.g. servers, storage, applications, services, and so on). The essential characteristics of cloud computing include on-demand selfservice, broadband network access, resource pooling, rapid elasticity, and measured service [1]. Several service models are supported, including cloud software as a service, cloud platform as a service, and cloud infrastructure as a service. Cloud computing has attracted significant attention, and several commercial clouds, including Amazon EC2, Microsoft Azure, and Google App Engine, have been providing services to users.Cloud computing will have profound impacts on the design and operation of next generation mobile wireless cellular networks. On the one hand, with advances of wireless mobile communication technologies and devices, more and more end users access cloud computing systems via mobile devices such as smart phones and tablets. The integration of cloud computing into the mobile environment enables mobile cloud computing (MCC), which is widely considered as a promising mobile computing paradigm with a huge market [2]. MCC enables offloading the computing power and data storage requirements from mobile devices into the powerful computing platforms in the cloud, bridging the gap between the increasing computing demands and the traditional mobile computing technologies with limited computing, storage, and energy resources in mobile devices.On the other hand, the powerful cloud computing technology can be beneficial to radio access networks (RAN) as well (in addition to mobile end users), which leads to a novel concept of cloud radio access networks (C-RAN) [3,4]. Unlike the existing cellular networks, where computing resources for baseband processing are located at each cell site, in C-RAN the computing resources are located in a central wireless network cloud with a powerful computation resource pool. This transition from distributed to centralized infrastructure for baseband processing can have significant benefits: saving the operating expenses due to centralized maintenance; enabling better load balancing; improving network performance due to advanced coordinated signal processing techniques; and reducing energy expenditure by exploiting the load variations.Although some excellent work has been done to study cloud computing for both end users and access networks, these two important areas have traditionally been addressed separately in the literature. To the best of our knowledge, the joint study of C-RAN in MCC for next generation cellular networks has not been addressed in previous work. In this article we study the topology configuration and rate allocation problem in C-RAN with the objective of optimizing the end-to-end TCP throughput performance of MCC users in next generation cellular networks. Despite the potential benefits brought by C-RAN, one of the major challenges in C-RAN is that t...
