Abstract

In order to provide better Quality-of-Service (QoS) in large networks, several congestion pricing proposals have been made in the last decade. Usually, however, those proposals studied optimal strategies and did not focus on implementation issues. Our main contribution in this paper is to address implementation issues for congestion-sensitive pricing over a single domain of the differentiated-services (diff-serv) architecture of the Internet. We propose a new congestion-sensitive pricing framework Distributed Dynamic Capacity Contracting (Distributed-DCC), which is able to provide a range of fairness (e.g. max-min, proportional) in rate allocation by using pricing as a tool. Within the Distributed-DCC framework, we develop an Edge-to-Edge Pricing Scheme (EEP) and present simulation experiments of it. Document type: Part of book or chapter of book

Abstract

Most congestion control algorithms try to emulate processor sharing (PS) by giving each competing flow an equal share of a bottleneck link. This approach leads to fairness, and prevents long flows from hogging resources. For example, if a set of flows with the same round trip time share a bottleneck link, TCP’s congestion control mechanism tries to achieve PS; so do most of the proposed alternatives, such as eXplicit Control Protocol (XCP). But although they emulate PS well in a static scenario when all flows are long-lived, they do not come close to PS when new flows arrive randomly and have a finite amount of data to send, as is the case in today’s Internet. Typically, flows take an order of magnitude longer to complete with TCP or XCP than with PS, suggesting large room for improvement. And so in this paper, we explore how a new congestion control algorithm — Rate Control Protocol (RCP) — comes much closer to emulating PS over a broad range of operating conditions. In RCP, a router assigns a single rate to all flows that pass through it. The router does not keep flow-state, and does no per-packet calculations. Yet we are able to show that under a wide range of traffic characteristics and network conditions, RCP’s performance is very close to ideal processor sharing. Document type: Part of book or chapter of book

Abstract

Traffic engineering tools are applied to design a set of paths, e.g., using MPLS, in the network in order to achieve global network utilization. Usually, paths are guaranteed long-term traffic rates, while the short-term rates of bursty traffic are not guaranteed. The resource allocation scheme, suggested in this paper, handles bursts based on maximal traffic volume allocation (termed TVAfB) instead of a single maximal or sustained rate allocation. This translates to better SLAs to the network customers, namely SLAs with higher traffic peaks, that guarantees burst non-dropping. Given a set of paths and bandwidth allocation along them, the suggested algorithm finds a special collection of bottleneck links, which we term the first cut, as the optimal buffering location for bursts. In these locations, the buffers act as an additional resource to improve the network short-term behavior, allowing traffic to take advantage of the under-used resources at the links that precede and follow the bottleneck links. The algorithm was implemented in MATLAB. The resulted provisioning parameters were simulated using NS-2 to demonstrate the effectiveness of the proposed scheme. Document type: Part of book or chapter of book

Abstract

This paper focuses on the selection of bandwidth-guaranteed channels for communication sessions that require it. The basic idea comes from Minimum Interference Routing: select a feasible path that puts the least possible restriction on the transmission capacity offered by the network for other communicating parties. This is achieved by circumventing some critical bottleneck links. The main contribution of the paper is a method to assess the degree of link criticality facilitating efficient route precomputation even in the case, when up to date resource availability information is not immediately available. Document type: Part of book or chapter of book

Abstract

MaxNet is a distributed congestion control architecture in which only the most severely bottlenecked link on the end-to-end path controls the source rate. This paper shows that the small-signal convergence speed of MaxNet is higher than that of conventional architectures, such as the current Internet or REM. It also shows that MaxNet decouples the control, so that each pole position depends only on parameters of one bottleneck link and of the sources controlled by that bottleneck, enabling optimal pole placement. Document type: Part of book or chapter of book

Abstract

This paper considers a congested bottleneck. A fast lane reserves a more than proportional share of capacity to a designated group of travellers. Travellers are otherwise identical and other travellers can use the reserved capacity when it would otherwise be idle. The paper shows that such a fast lane is always Pareto improving under Nash equilibrium in arrival times at the bottleneck and inelastic demand. It can replicate the arrival schedule and queueing outcomes of a toll that optimally charges a constant toll during part of the demand peak. Within some bounds, the fast lane scheme is still welfare improving when demand is elastic. Document type: Book

Abstract

Abstract: "The ability to locate network bottlenecks along end-to-end paths on the Internet is of great interest to both network operators and researchers. For example, knowing where bottleneck links are, network operators can apply traffic engineering either at the interdomain or intradomain level to improve routing. Existing bandwith measurement tools fail to identify the location of bottleneck links. In addition, they often require access to both end points and generate huge amount [sic] of probing packets. These drawbacks make them impractical. In this paper, we present a novel light-weight, single-end active probing tool -- Pathneck -- based [sic] a novel probing technique called Recursive Packet Train (RPT), which allows end users to efficiently and accurately locate bottleneck points to destinations on the Internet. We evaluate Pathneck using trace-driven emulations and wide area Internet experiments. In addition, we conduct extensive measurements on the Internet among carefully selected, geographically diverse probing sources and destinations to study Internet bottleneck properties. We find that Pathneck can successfully detect bottlenecks for over 70% of paths, and most of the bottlenecks are fairly stable. We also report our success on bottleneck inference, using multihoming and overlay routing to avoid bottlenecks based on the bottleneck link location and bandwidth estimation provided by Pathneck." Document type: Report

Abstract

This discussion paper led to a publication in 'Transportation Research Part B: Methodological' . The traditional bottleneck model for road congestion promotes the implementation of a triangular, fully time varying, charge as the optimal solution for the road congestion externality. However, cognitive and technological barriers put a practical limit to the degree of differentiation real world implementations can handle. The traditional approach to accommodate for this concern has been a step toll, with the single step coarse charge as its simplest case. In this paper we study how efficiency of the coarse charge can be improved by differentiating its level and timing across groups of travellers. We use the traditional bottleneck model to analyse how the coarse charge can be differentiated over two groups of travellers assuming inelastic peak-hour demand. The results of our analysis indicate that differentiating the coarse charge across two groups of travellers considerably improves its efficiency without increasing cognitive effort and decision making costs for the individual traveller. A numeric illustration reveals a welfare gain of 69% of the first best charge, up from 53% for the generic coarse charge. This increase is similar to what is obtained by moving from the coarse charge to a generic two step toll. Once different groups have been defined, one could in fact achieve the same gains by temporal separation of drivers, for example by use of licence plate numbers. The presented charging regime has a considerable degree of flexibility with respect to share of travellers to attribute to each scheme, which further adds to its merits in practical applicability.

Abstract

This discussion paper resulted in a publication in 'Transportation Research Part B: Methodological' , 2013, 53, 31-44. We analyse the behaviour of market participants in a multi-modal commuter network where roads are not priced, but public transport has a usage fee, which is set while taking the effects on the roads into account. In particular, we analyse the difference between markets with a monopolistic public transport operator, which operates all public transport links, and markets in which separate operators own each public transport link. To do so, we consider a simple transport network consisting of two serial segments and two parallel congestible modes of transport. We obtain a reduced form of the public transport operator's optimal fare setting problem and show that, even if the total travel demand is inelastic, serial Bertrand-Nash competition on the public transport links leads to different fares than a serial monopoly; a result not observed in a static model. This results from the fact that trip timing decisions, and therefore the generalized prices of all commuters, are influenced by all fares in the network. We then use numerical simulations to show that, contrary to the results obtained in classic studies on vertical competition, monopolistic fares are not always higher than duopolistic fares; the opposite can also occur. We also explore how different parameters influence the price differential, and how this affects welfare.

Abstract

Discussions of competition and regulatory reform typically focus on price and quantity effects. But improving certain infrastructure services can also stimulate entry, and competition in user industries downstream, allowing new firms to enter, incumbent users to offer new products, and rivalry to intensify. The authors present a case study of how innovations in road freight services affect selected downstream users of those services after regulatory reform. After a period of rigid regulation, and heavy government interference, Mexico in 1989 developed a new policy framework for road transport, with free entry, and market-based price setting. The result: faster, more reliable trucking has allowed user companies to offer new, previously unavailable products, and to reach new areas with existing products. Cheaper, more customer-responsive trucking services have allowed logistical innovations in user firms, and some user firms have decided not to keep their own fleets of trucks, but to outsource trucking services on the open market, thereby converting fixed costs to variable costs. For one fertilizer company, the benefits of reform included a ten percent improvement in operating margin. Successful reform requires careful planning and execution, and political support at high levels. Regulatory reform also profoundly changes the sectoral institution formerly responsible for the regulation. Enough resources should be provided to help organizations in the reformed industry make the transition to the post-reform environment - helping with such tasks as defining the organizations new role, and facilitating the redeployment of staff. The national competition agency can help greatly in laying the groundwork for reform by making a compelling case for the reforms expected benefits. After reform, the competition agency should also help with enforcement, to ensure that the cozy, cartel-like behavior stimulated by tight entry restrictions does not persist. In Mexico, three strong interventions were required to discipline attempted anti-competitive practices in the trucking industry in the years following reform. Document type: Book