Impact of the Net Neutrality Repeal on Communication Networks: Related Works

Table of Links

Abstract & Introduction

Related Works

Repealing Net Neutrality

Profit-Driven Model Results

Conclusions & References

Biographies

II. RELATED WORKS

Many papers in the literature discussed and analyzed various aspects of net neutrality. From a legalization and regulation perspective, net neutrality in the Internet ecosystem has been surveyed by the authors in [14] and [15]. They emphasized that cloud computing has initiated the net neutrality battle between ISPs and CPs. In [16] the authors analyzed the Internet video streaming contest, taking into account all of ISPs and CPs assets (e.g. content rights, network access, users, …etc). They stated that video distribution makes the dilemmas of net neutrality solid and perceptible. Their analysis demonstrates that net neutrality correlates highly with video service delivery at different points including competition between CPs and ISPs, competition between stand-alone CP and CP owned by ISPs in providing video delivery services and growth of video traffic.

A number of papers in the literature focus on providing mathematical models to investigate the influence of the repeal of net neutrality on communication networks. Paid service differentiation where CPs voluntarily pay a monopoly ISP for prioritizing their traffic under shared network infrastructure was investigated by the authors in [17]. The differentiation occurs where ISPs offer service classes for CPs to choose from where traffic of a higher-priority class will be processed before those of a lower-priority. They studied the optimal pricing based on either maximizing the CPs’ choices of service classes or minimizing system delays. Consequently, they highlighted that ISPs optimal pricing strategy can result in an efficient differentiation among CPs maximizing social welfare. Also, they found that applying paid prioritization can lead to money flows (profit) from CPs to ISPs. The authors in [9] modelled the competition of video services delivery market between an ISP’s own integrated CP and stand-alone CP. They studied the impact of applying different QoS (marking video traffic as high priority) pricing strategies either by selling QoS to CPs, selling QoS to users, or choosing to not provide QoS at all. They investigated the impact of QoS pricing on the video service prices and CPs profit. The analysis showed that ISPs can sell QoS to CPs at a higher price than when QoS is sold to users, and the CPs are able to make more profit when QoS is directly sold to users than the case when QoS is sold to CPs. Also, they found that an ISP is more likely to use QoS exclusively for its own video services when it provides a similar content of CPs. The cloud infrastructure needed to host and deliver the video content was optimized in [18]-[21] and the impact of the delivery of large data volumes on the network was evaluated in [22]-[25]. Particular attention was paid to the core network which forms the heart of the ISP infrastructure and hosts the CDN with attention given to the network energy efficiency, latency and other QoS metrics [26]-[33]. The work in [34] considered the impact of maximizing profit of CDN providers considering users who access CPs content from either cloud or fog server. In the case of competitive CPs, the CDN always places the content of the popular CP in fog servers, even when a less popular CP pays more, as the CDN tries to reduce core network transit cost.

In this paper, a techno-economic Mixed Integer Linear Programming (MILP) model is built to study the potential profit ISPs can achieve by a differentiated pricing scheme under the repeal of net neutrality. We build on our MILP optimization, network, cloud and fog modelling background [35]-[39] and consider ISPs that offer the CP service classes, which represent different data rate requirements. The model optimizes the pricing scheme of differentiated service classes to maximize the ISP profit based on price elasticity of demand (PED). The MILP model finds the resulting equilibrium pricing, core network power consumption and traffic.