IEEE Consumer Electronics Magazine - March/April 2021 - 86

Consumer Technology-Based Solutions for COVID-19

Table 2. General simulation parameters.
Parameter

Values

Latency between UE and the MEC IBS: tUE !MEC

0.25 ms

Latency between UE and the AN: tUE!AN

1 ms

Latency between AN and the CN: tAN !CN

0.05ms/
km24

AR processing delay for UC 1 and UC 3: tAR

30 ms24

Audio Video processing delay for UC 2: tAV

20 ms

Distance to the RAS control station in UC 3: dAR

50 km

Bandwidth of the backhaul network BWAN !CN

10 Gb/s

Bandwidth requirement for UC 1: BWUC1

24 Gb/s

Figure 3. Model for conducting the simulations.

Bandwidth requirement for UC 2: BWUC2

2 Gb/s

Bandwidth requirement for UC 3: BWUC3

30 Gb/s

Furthermore, AR-based RAS control station is
located dAR distance in the same direction
toward the CN. The latency associated toward
AN to CN, UE to AN, and UE to MEC IBS are
denoted as tAN !CN , tUE!AN , and tUE!MEC ,
respectively. Moreover, we assume that the corresponding devices and processes related to
each use case are operating independently during our simulations. The parameters specified in
Table 2 are used for conducting the evaluations.

Computing capacity of MEC edge: CPU cycles
Computing capacity of the cloud platform: CPU
cycles

5 Â 1010 16
1011 16

system (such as PHANToM)23 should be implemented with the surgical controlling function.

FEASIBILITY EVALUATION
We consider two evaluation scenarios to validate our claim of requiring an edge computing
platform to realize the above discussed use
cases. The results are simulated for latency and
scalability aspects for MEC edge scenario considering all 3 Use Cases (UCs), in comparison to
an operating environment with cloud computing
capabilities. Furthermore, a forecasted expenditure analysis is presented at the end of this
section.
Figure 3 illustrates the model we have considered for conducting the simulations. For the
MEC enabled scenario, UE is directly accessing
the resources in the edge infrastructure at the
IBS in a closer range. In the cloud computing circumstance, UE is reaching the cloud environment or CN through the access network (AN)
formed by an eNB displaced 1 km from the hospital premises. This eNB is connected to the MEC
IBS for catering external connections to the MEC
edge. We assume that the link between AN and
the CN is established via a fiber optic connection
of BWAN !CN bandwidth spanning to dCN kms.

Impact on Latency
In this scenario, end-to-end latency is compared for MEC enabled and disabled instances. Under UCs 1 and 3, AR processes were
assumed to be latency critical applications,
where the delay associated with processing
an AR stream was computed. In UC 3, it is
assumed that the RAS control station is
located 50 km away from the surgical theatre.
For MEC enabled scenario, latency is computed as a combination of tUE!MEC and tAR ,
tAV or tMEC !RAS in UC 3. In contrast, for cloud
scenario, latency is computed by aggregating
tAN , and distance dependent tAN !CN apart
from processing delays. Thus, Figure 4(a) indicates the latency associated for each UC for
MEC enabled scenario (i.e., highlighted); in
contrast to CNs located at different distances
ranging from 50 to 500 kms. The latency for
UC 3 is showing a better result for CNs in
close proximity due to the external access of
RAS control sequence. Other results however,
indicate that MEC based IBS is satisfying the
service requirements specified in Table 1.

IEEE Consumer Electronics Magazine

IEEE Consumer Electronics Magazine - March/April 2021

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - March/April 2021