Solving the capacity and data rate challenges with network densification could be very expensive in terms of equipment, maintenance, and operations in future 5G. In LTE and LTE-A most of the functionality of RAN and L2 layer was put on e-Node B. For 5G again it is being speculated to make simpler BS's. What would be the real solution to this?
One way to reduce equipment cost is to minimize the number of functionalities at the
base station. This could be done by implementing only layer 1/2 (L1/L2) functionalities in the base station and moving higher-layer functionalities to a network cloud that serves many base stations. Reducing the number of functionalities results in simpler base stations, which could be deployed by users and remotely or autonomously managed to reduce deployment and operation costs.
From my practical experience in cellular networks deployments, the major cost is associated with the installation and with the services (man-power) required to put the base station on air, also with the backhualing expense. The price of the equipments is a minor contributor to the total cost (whether L2 or L3). Moreover this price is getting lower by time due to the advancements in computation power and reduction in electronics cost.
This fact explains why relay nodes have not been adopted by commercial cellular operators in spite the standardization effort in 3GPP.
Accordingly, very-small cells are still not widely adopted telecom operators, due to the lack of the proper business case, where the vast majority of traffic is still generated from conventional base stations.
In dense networks, the cloud computing capability of the back-haul network can be exploited to make very simple transmission techniques at base stations feasible in order to effectively manage the interference and enhance network throughput. Commercialization of this approach has been reported as 'pCell' technology:
http://spectrum.ieee.org/telecom/wireless/can-artemis-deliver-5g-service-on-your-4g-phone
Cloudification of radio access networks might solve issue to an extent. However, as Akram mentioned, pricing and business case remains an open ended problem from operator perspective.
- Badges
- Science topic