Page 224 - Proceedings of the 2017 ITU Kaleidoscope

S4.4 Drone readiness index
Samuel Nzaramba (Carnegie Mellon University Africa, Kigali, Rwanda), Rene Kabagamba
(Carnegie Mellon University Africa, Kigali, Rwanda), Kate Chandler (Georgetown University,
Washington DC, USA), Aminata Garba (Carnegie Mellon University Africa, Kigali, Rwanda and
Carnegie Mellon University, Pittsburgh, USA)

This paper proposes a new model for evaluating the robustness of the ecosystem for drone
projects in a given country, considering nine factors ranging from the regulatory framework to
economic and social impact. The objective of this study is to provide a tool in the form of an
index that can be used to gauge countries readiness for drone projects. Governments, NGOs as
well as commercial drone companies can use the index to gain insights into the possibilities of
drones for non-military use. Notable successful projects using drones were used as a benchmark
to chart out the various components of the Drone Readiness Index (DRI). We first reviewed
selected projects that have attempted to use drone aircrafts for non-military activities, using
secondary data. We then quantify the elements of the drone ecosystem and present derivations of
the proposed drone readiness index. To show applications and examples of the proposed drone
readiness index, we compute the values of the drone readiness index for selected African
countries. These values are further presented in a website.

Session 5: Smartening up society with data and new applications
Machine learning approach for quality adaptation of streaming video through 4G wireless
S5.1
network over HTTP
Dhananjay Kumar, Aswini Viswanathan (Anna University, MIT Campus, India); Arun Raj
Lakshminarayanan (B.S.A Crescent University, Chennai, India); Hiran Kumar Singh (Vel Tech
University, Chennai, India)

Video streaming over HTTP through 4G wireless network used for multimedia applications
faces many challenges due to fluctuations in network conditions. The existing HTTP Adaptive
Streaming (HAS) techniques based on prediction of buffer state or link bandwidth offer solution
to some extent, but if the link condition deteriorates, the adaptation process may reduce the
streaming bit rate below an acceptable quality level. In this paper, we propose a machine
learning based method, State Action Reward State Action (SARSA) Based Quality Adaptation
algorithm using Softmax Policy (SBQA-SP), which identifies the current state (Throughput),
action (Streaming quality) and reward (current video quality) at client to determine the future
state and action of the system. The ITU-T G.1070 recommendation (parametric) model is
embedded in the SBQA-SP to implement adaptation process. The proposed system was
implemented on the top of HTTP in a typical internet environment using 4G wireless network
and the streaming quality is analyzed using several full reference video metrics. The test results
outperformed the existing Q-Learning based video quality adaptation (QBQA) algorithm. For
instance, an improvement of 5% in average PSNR and 2 % increase in average SSIM index over
the QBQA approach was observed for the live stream.

– 208 –