Abstract:
Proliferating use of mobile communications have urged the need to develop networks able to cater
to higher bandwidth, speeds and support a plethora of upcoming technologies. Introduction of 5G
networks in a heterogeneous network architecture has been chosen as a viable solution to persistent
issues in current implementations. However, these network designs lack several fundamental
software and hardware pitfalls associated with problems in designing: associated cell optimizations,
schemes on simultaneous base station associations and cooperation between tiers in the
architecture. Therefore, this research will focus in fine tuning these software and hardware fallacies
for the successful implementation of proposed 5G networks.
A main software drawback in current networks is persistence of lazy caching themes. As shown
in figure 1, currently user requests are often matched to arbitrary locations without the use of preenabled
caching mechanisms. To overcome this issue proactive caching where base stations (BS)
identify external clients possessing cached information and dynamic Device-to-Device (D2D)
connection creation could be implemented. As shown in figure 2, significant improvements in
successful requests could be achieved both under high load and under low load as users are
efficiently matched to potential targets.
A main drawback in mobile network hardware design is high energy consumption proportional to
increasing user requests. To overcome this issue cell zooming could be introduced to 5G
implementations. As shown in figure 3, the central cell could use an algorithmic approach to
identify the network request density around it and zoom its range in or out to efficiently serve
while neighboring cells could be switched off for a predetermined interval and vice versa. T-tests
carried under this paradigm proved that significant cost savings in efficient use of energy in these
cells could be achieved under this solution
