Article Preview
TopIntroduction
Cloud Radio Access Networks (C-RAN) offers a promising solution to the high data and coverage demand of the users with the passage of time by leveraging the computing capabilities of the cloud platform. It is one of the key technology used in 5G networks being much more efficient as compared to the existing heterogeneous architectures (Damnjanovic et al., 2011; Jimma, Chai, Chen, & Alfadhl, 2011). As the existing cellular networks have large capital expenditure (CAPEX) and operational expenditure (OPEX) and long installation times associated with it, C-RAN becomes a key technology in this regard as being a much more resource and cost-efficient technology. C-RAN makes use of centralized base band processing through a centralized pool called as base band unit (BBU) (Checko et al., 2011). A C-RAN architecture consists of remote radio heads (RRH) serving as radio transmission antennas, connected through a high-speed link to the data center. In the data center also called as base band unit, there are different protocol layers such as Physical layer etc. virtualized in a software on the centralized cloud computing platform (Wu, Zhang, Hong, & Wen, 2015). This architecture ensures efficient resource utilization, reduced power consumption and easy upgradable/reconfigurable architecture. The general model of the Cloud radio access network is shown in the Figure 1.
C-RAN concept is also closely related to software defined networking, the major objective of this technology is to provide support to distribute computing compatibilities and also to allow different network operators to share physical infrastructure (Dahrouj, Douik, & Dhifallah, 2015; Peng, Li, Zhao, & Wang, 2015). Cloud radio access network consists of distributed processing units, resource allocation and coordination between these units require a centralized scheduler (Soliman & Leon-Garcia, 2016). The number of users are increasing with the passage of time, the number of RRHs have also been increased. This increase in number of users/RRHs, higher bandwidth and higher physical layer complexity may result in congestion of scheduler performance. As the communication that takes place between the scheduler and processing units is very large in a very short time duration e.g. less than 1 ms is being predicted for 5G. Thus, scheduler plays a very significant role in evaluating the efficiency of the network. Time varying channel conditions are very important feature of the wireless networks, and are independent for different users. This feature is utilized by the opportunistic scheduling methods to get the multi-user diversity gain and also increase the throughput of the system. However, it is also very important to maintain fairness among all the users while providing an acceptable data rate to them.
In wireless communications, the channel conditions are variable with the time because of the shadowing and the fading effect. Thus, at a given time slot users experience different channel conditions. This results in multi user diversity effect, which states that if many users fade independently at a particular time slot then the probability that some users will have strong channel than the others is very high. If the users with strong channel conditions are allowed to transmit only, this leads to the most efficient utilization of the shared channel and the total throughput of the system is maximized. These type of scheduling mechanisms are also called as opportunistic mechanisms because they make full use of the favorable channel conditions. These type of scheduling mechanisms can offer several advantages such as higher system throughput and spectrum utilization. However, before realizing these advantages several other aspects should also be taken into account.