Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 48977 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Hossein Khalaj, Babak
- Abstract:
- There has been a continuous increase in requested data rates in both uplink and downlink of wireless communications for a long time. But in recent years, this increase has become more extensive due to widespread use of smart phones and social networks. In order to cope with such a heavy demand, wireless telecommunication operators, have to improve frequency efficiency, by implementing more and more base-stations all over the network. But base-stations are very expensive instrument, due to their complex structure and having a high processing power. Therefore, building a network with so many base stations, may be as expensive that operators cannot afford it. “Cloud Radio Access Network”, or CRAN, is a survival solution for the aforementioned problem. In CRAN structure the functionalities of a custom base-station is divided into two parts, “RF functionalities” and “processing functionalities”. The former has to be spread through the network, but the latter can be integrated into a “central processor”. So, the new base-stations are actually simple and cheap antenna, which makes implementation of numerous base-stations, economic. Integration of processing operations from all the base-stations, gives the capability of “joint processing to mitigate intra-cellular interference”, “parallel processing” and “computational load balancing on processors”. But there is a challenge and is that we have to transfer radio signals between the “central processor” and base-stations for both the uplink and downlink of the communications and that the capacity between base-stations and the central processor is limited. Thus one major problem in CRAN concept, is how to optimally allocate this limited fronthaul capacity to base-stations considering some specific quality of service (QoS) criteria. One important QoS criteria is minimizing the delay and its variations (jitter), that users experience in applications such as video call, video streaming, or video on demand. The suitable framework to consider users’ delay in a communicational network is Markov Decision Process (MDP). A recent research has conducted a method for allocation of the fronthaul capacity to the base-stations, in order to minimize the average sum delay of the users. This method, assumes that the wireless channel matrix is strongly diagonal and replaces the constraint “limited fronthaul capacity”, with the constraint of “limited average use of fronthaul capacity”. In this thesis we propose a method for allocation of backhaul capacity to minimize the average sum delay of users, with real channel matrixes and considering the real constraint of “limited fronthaul capacity”.Another important QoS criteria, is the efficient use of energy or the battery charge of users’ equipment. In this thesis, we also propose a method for jointly optimal “backhaul capacity allocation” and “control of the users’ transmit power” with the goal of minimizing the average sum delay of users.Finally, we consider the probability of finishing the service of a user or some, and solve the optimal backhaul allocation problem for minimizing the sum delay of users, by modeling the problem, as a “Total Discounted Reward” MDP problem
- Keywords:
- Service Quality ; BASE STATION CONTROLLER ; Markov Decision Making ; Cloud Radio Access Network ; Fronthaul Capacity ; Frequency Efficiency
Digital Object List
محتواي کتاب
Bookmark