J. MacKie-Mason and H. Varian. Pricing the Internet. In Public Access to the Internet (Kahin B. and Keller J., Eds.), Prentice Hall 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....congestion the price should reflect the incremental social cost determined by the marginal delay cost to other users and the willingness of the user to pay for the cost. Price adjusting can be performed by auction pricing and by feedback pricing. With auction pricing or a smart market approach [25], prices are determined based on consumers bids. Users include a bid in each packet, or based on a larger granularity [26] At congested routers, packets are prioritized based on these bids. In case of congestion, packets of users offering the lowest bid are discarded first, and accepted packets ....

....The cost of carrying each packet is thus related to the marginal value (represented by the bid) of the traffic which has been pushed out. At the equilibrium price the user s willingness to pay for additional data packets equals the marginal increase in delay cost generated by those packets [25]. Auctioning of bandwidth, rather than auctioning based on individual packets, has been considered in [16[23] In this approach, bandwidth is split into small units, and users bid for the required bandwidth at each auction period. With feedback pricing, the prices are calculated by the provider ....

J. MacKie-Mason and H. Varian. Pricing the Internet. In Public Access to the Internet (Kahin B. and Keller J., Eds.), Prentice Hall 1995.

....in which each application and user will require a different QoS, pricing network traffic based on usage will be a necessity, perhaps combined with fixed access charges, for recovering costs. Pricing Internet Traffic There have been several pricing proposals in recent literature (see, for example [2, 5, 7, 11]) A comprehensive review of these pricing proposals is presented in [6, 12] The basic idea in economic pricing of network traffic is to charge the incremental cost of providing passage. Since the short term incremental cost of providing passage through fixed capacity computer networks is ....

MacKie-Mason, J., and Varian, H. Pricing the Internet. In B. Kahin and J. Keller, Eds., Public Access to the Internet. Prentice-Hall, NJ, 1995.

....market, suggesting the need for flexibility in contract establishment by means of performing reservations. We consider this a promising direction for further work. We are also interested in the issue of pricing. In the current version of our prototype, prices are fixed, rather than usage based [MMVa95]. An important question remains regarding usage based pricing schemes for composite services, involving more than one resource. CoSiSt97] and [CoSi98] present results concerning usage based pricing for bandwidth in ATM networks, based on simple measurements and the concept of effective bandwidth, ....

J.K. Mackie-Mason and H.R.Varian. "Pricing the Internet". In B. Kahin and J. Keller, editors, "Public Access to the Internet", Prentice Hall, 1995.

....locally operating agents can achieve this in an efficient manner. In the text, we discuss the case of tolling on a specific road segment upstream of the bottleneck. This demands prior knowledge about the system. However, one can imagine a completely local algorithm in the following way (see also [40]) Assume that every road segment in the system is operated by a simple economic agent. This agent wants to keep the operation of the segment as efficient as possible, and the only measure she has is to go up or down with the toll. The agent knows the performance characteristics (i.e. throughput ....

MacKie-Mason, J.K., and H.R. Varian, Pricing the internet, preprint 1994.

....or for delivering service to those users who need it most. Pricing mechanisms have the potential of overcoming these shortcomings, resulting in more efficient resource allocation, by charging users on the basis of the congestion that they cause. In this spirit, several models have been put forth [3, 4, 5, 6]. For a more detailed literature review see [1] The emergence of real time traffic substantially complicates the picture and requires QoS measures much harder to analyze (see, e.g. 7, 8] A central question that has received little formal treatment relates to determining the right time scale ....

J. K. MacKie-Mason and H. R. Varian, "Pricing the Internet," in Public Access to the Internet (B. Kahin and J. Keller, eds.), Prentice-Hall, 1994.

....a smaller fraction of the usable storage space than the meta data space required for smaller objects. The utility of network caching becomes clear through straightforward cost analysis considerations. The cost of internet class bandwidth has been reported to be decreasing at a rate of 30 per year [5]. Hard disk prices are also decreasing at an exponential rate [6] However, in the case of disk prices, the rate is 50 per year. Considering the current cost of bandwidth and disk prices, we see that caching an object is beneficial if it is accessed just a handful of times per month [7] ....

Jeffrey K. MacKie-Mason and Hal R. Varian, "Pricing the Internet," Public Access to the Internet, May 1993.

....by the buffer length at a local gateway the longer the buffer, the higher the price. The price is fixed for a predetermined period, after which the buffer is sampled again and a new price is assigned, if necessary. An alternative scheme, called smart market pricing, was introduced in [20]. In this proposal, users bid for network access on a per packet basis by telling their local gateway (in each packet s header) how much they are willing to pay for the packet to get on to the network. The gateway admits the packets with the highest bids, but all users whose packets have been ....

....current Internet [5] Proponents of usage based pricing claim that this discipline will reduce congestion by making users pay for the network capacity that they use. In particular, they argue that usage based pricing will enable delay sensitive applications such as telephony and video conferencing [20]. However, the advantages of usage based pricing do not come for free there will be significant difficulties implementing such a scheme in the current Internet. In particular, multimedia users will find that they will have to pay orders of magnitude more for a video conference as opposed to ....

J. K. MacKie-Mason and H. Varian, "Pricing the Internet," in Public Access to the Internet, ed. B. Kahin and J. Keller, Prentice-Hall, 1995.

....end users. The provision of integrated services to many end users (with inherently very different demands for transmission quality) while avoiding the tragedy of commons (Hardin 1968, Gupta et al. 1995) will raise the issue of charging users according to their actual use of resources (MacKie Mason and Varian 1993). The connectionless Internet architecture has proven to be extremely robust and suitable for data traffic in heterogeneous environments. It is used by millions of users every day. The pricing mechanism proposed in this paper was developed with the current Internet in mind. It is focused on ....

MacKie-Mason, J. K. and Varian, H. R. (1993). Pricing the Internet. In "Public Access to the Internet". JFK School of Government, Harvard University.

....to support mixed traffic profiles, including real time continuous media traffic, with orders of magnitude more bandwidth capacity, complex resource reservation, accounting and pricing schemes. However these proposals may well require major redesign, preceded by considerable discussion [2] 6] 8] [7] [9] In the interim we have a time window in which the Internet must accommodate traffic for which it was not designed. 3. Hope that service providers will be able to continue to upgrade the network ahead of demand, with no transference of resource consumption and upgrade costs to users. Never ....

J. Mackie-Mason and H. Varian. Pricing the Internet. Technical report, U. Michigan, May 1993.

....use of priority service based on application requirements and usage policies. While pricing is not the focus of this study, we note that one of the key advantages of the proposed architecture is that it can provide simple mechanisms for calculating nearoptimal prices based on congestion costs [15]. In the following sections, we demonstrate the efficacy and the robustness of the proposed framework. Using extensive simulations, we show that the proposed architecture adapts with the traffic dynamics in the Internet to eliminate the risk of congestion collapse. When used in conjunction with ....

J. MacKie-Mason and H. Varian. Pricing the Internet. In Public Access to the Internet, pages 269--314, May 1995.

....outline one possible method for computing adaptive prices. The users discussed below are thus adaptive users ( either adaptive best effort, inelastic, or elastic ) A more detailed description can be found in [11] 12] Other pricing schemes for communication networks have been suggested, e.g. [13], 14] The network and its users are considered to form an economy or economic system. The system has various resources such as link bandwidths and buffer spaces that can be used to meet user demands for service. Network constraints such as buffer sizes or link capacities are translated into cost ....

J. K. MacKie-Mason and H. R. Varian, `Pricing the Internet', presented at Public Access to the Internet, JFK School of Government, Harvard University, May 26-27, 1993.

....goals across organisations, it appears likely that money will be used to justify decisions, and the network will have a pricing scheme designed to control congestion. An obvious policy is to charge users more in periods of congestion, such as the smart auction suggested by Mackie Mason and Varian [72,73], where each user bids a price for their traffic to get through, the bids and their associated bandwidth requirements are ordered and the price chosen is the bid in the list where the bandwidth demands of all higher prices sum to the available bandwidth. All users who are prepared to pay a higher ....

Jeff Mackie-Mason & Hal Varian, "Pricing the Internet," Proceedings of Public Access to the Internet, Cambridge, Ma (1993).

....as phone companies do today, such as flat fees, or flat up to some usage point, with usage based billing beyond that point. Although increasingly critical, these topics are beyond the scope of this thesis. Several studies have investigated pricing based on usage and or priority [117] 118] 119] [120]. However, prerequisite to accounting and billing instrumentation is a more accurate model for the attribution of resource consumption, including distinguishing among different qualities of service and between sender initiated (e.g. e mail) and receiverinitiated (e.g. gopher, mosaic, www, ftp) ....

J. MacKie-Mason and H. Varian, "Pricing the Internet," in Public Access to the Internet (B. Kahin and J. Keller, eds.), Prentice-Hall, 1994. Available from ftp://gopher.econ.lsa.umich.edu/pub/Papers.

....one must consider user incentives to ensure that these QoS features are used appropriately. Appropriate use leads to the efficient use of network resources, where efficiency is defined in [4, 20] These incentive issues are also relevant to controlling usage even in a single QoS environment (see [8, 9]) 2 We should also point out that even though costs are allocated to individual users, the recovery of those costs might occur at a much higher level of granularity. For instance, USC might be responsible for all of the costs allocated to its students and staff. 3 Calculating the costs for a ....

....approach to satisfy the axioms presented in Section 3. We conclude in Section 5 with a discussion of various directions for future work. Before turning to the basic model, we first discuss some related work. There is a rapidly growing literature on pricing in computer networks. See References [8, 4, 17, 2, 20] for a few representative examples. However, this literature does not typically relate the prices charged to any underlying cost of network usage. The emphasis is on the strategic (utility maximizing) behavior of users, and the profit maximizing behavior of firms, rather than on the equitable ....

J.K. MacKie-Mason and H.R. Varian. Pricing the Internet. In B. Kahin and J. Keller (Eds.), Public Access to the Internet. Prentice-Hall, Englewood Cliffs, New Jersey. Available from ftp://gopher.econ.lsa.umich.edu/pub/Papers, May 1993.

....rise to new problems that have not been addressed by previous research in real option theory. For example, prior research does not consider decommitment in contracts, contract quality guarantees, or sequential subcontracting. To put our research in context, we mention that the economic perspective [24, 48]is gaining popularity in computer science in general and in AI in particular. For example, 1 We use the word contingent or flexible contracts in our document to distinguish from the contingency contracts of game theory. In contingent contracts, the particular future contingenciesdo not need to ....

J. MacKie-Mason and H. Varian. Pricing the internet. In B. Kahin and J. Keller, editors, Public access to the internet. MIT Press, 1995.

....makes upon the network service providers. A network services market will be expected to offer multiple services at multiple rates, with low costs and latencies for switching from one service type to another. Market mechanisms capable of supporting these requirements are a topic of active research [9, 10]. As a second example, consider a market in which information brokers compete to provide information filtering services. As has been discussed, the varied preferences among users for different categories of information induce horizontal differentiation. However, there may be several vertical ....

J. K. Mackie-Mason and H. R. Varian, "Pricing the Internet, " Second International Conference on Telecommunication Systems Modelling and Analysis pp: 378393, Nasville, Tennesse, March, 1994.

....out a research agenda for economists in response to these changes in technology and lifestyle, which is partially based on my experience of working with computer scientists over the last several years. With the notable exceptions of research on congestion and pricing of networks (see, for example, Mackie Mason Varian (1995), or Hubermand Lukose (1997) it is disappointing how little work has been carried out by economists in response to these developments 3 . This is true not only because of the scale of these developments, but also because some of the issues raised are directly related to the types of problems ....

Mackie-Mason J.K., H.R. Varian (1995), "Pricing the Internet", In Kahin B., and Keller J. (Eds), Public Access to the Internet. MIT Press, Cambridge, MA.

....and global performance is similar to the one discussed above in the introduction: when every QA sends a query to the least loaded IA, everyone s performance suffers. Researchers taking the economic viewpoint have proposed that pricing internet access can lead to a resolution of this dilemma [12, 14, 17, 18]. Some researchers [10, 15, 19, 20] are pursuing the design of decentralized strategies using models based on market equilibrium [7] In this paper we are formulating the decentralized strategy design problem purely from a performance viewpoint: if each agent uses a strategy that leads to ....

J. MacKie-Mason and H. Varian. Pricing the internet. In B. Kahin and J. Keller, editors, Public access to the internet. MIT Press, 1995.

....will apply to each service class and hence each service class will be priced efficiently. The optimality approach in [Varian1] however suggets setting the price based on the marginal loss in utility of all the flows at a particular congestion level, which is quite acceptably, a hard problem. In [Varian2], they propose an approach that can alleviate this problem. This is the smart market approach based on the Vickrey Auction. Every packet in a flow carries a bid in its header which is supposed to reflect the valuation of the service by the user. During congestion in the network, the packets ....

....of previous and current prices) satisfies C(q(t) fi) C total , the total capacity of the network. For the range of service classes 1 to N , this constraint becomes C(q 1 (t) q N (t) fi 1 (t) fi N (t) C total For best effort service, the authors assume the bidding approach [Varian2], i.e. each packet carries a bid reflecting the valuation of their service and the network sets a current cut off price p b (t) which is a function of the current buffer occupancy and the bids. The best effort packets share a multiplexed buffer B s . If the packet s bid is greater than or equal ....

J.K.MacKie-Mason and H.R.Varian; Pricing the Internet. in Public Access to the Internet,Prentice-Hall, Englewood Cliffs, NJ; May 1993.

....equilibrium, because in this case, the length of the interval between price computations is zero and prices can be calculated only from the projected user demands. The bidding scheme can be used to reach an equilibrium fast but is rather unrealistic in the telecommunications market. Mackie Mason [MAC94b] and Kelly [KEL94] have also found that prices based on measures make truth telling desirable for the users. 3.3 Experimental Results We have constructed an example with a medium sized user population. We model the arrival rate l of every user in the population as a random variable. The first ....

J.K. Mackie-Mason and H.R. Varian, "Pricing the Internet", Proceedings of the Second International Conference on Telecommunication Systems Modelling and Analysis, pp. 378-393, Nashville, Tennessee, March 24-27, 1994.

....toward automated as well as human agents. Even in the current environment, economic issues are coming to the fore, as numerous proposals for internet billing protocols, transaction security, and electronic cash are being put forth for consideration. In some proposed schemes for internet pricing (MacKie Mason Varian 1994), prices and allocations are determined by dynamic bidding mechanisms. In such an environment, economically savvy computational agents will be at a premium. It is widely recognized that the information services network of the future is fertile ground for computational agents. We believe moreover ....

MacKie-Mason, J. K., and Varian, H. R. 1994. Pricing the internet. In Kahin, B., and Keller, J., eds., Public Access to the Internet. Prentice-Hall.

....platforms. Traditional centralised schemes [1] provide poor scalability, while large decentralised systems must cope with decision making based on incomplete data to remain scalable. To enable the required control over centralised and distributed resource scheduling proportionshare techniques [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] have been developed. These techniques replace the concept of simple abstract priorities with an ability to obtain resources. This ability is typically expressed by the concept of money. In such systems, resources are shared between clients 2 THE ERA SYSTEM 2 dependent on the amount of money ....

H. R. V. J. K. MacKie-Mason, "Pricing the internet," tech. rep., Department of Economics, University of Michigan, 1994.

....guarantee, and effective bandwidth ideas do not directly apply. Ideas for pricing this traffic are developing, and envisage prices which dynamically adjust to the level of congestion in the network. See Courcoubetis, Siris and Stamoulis (1998b) Kelly (1997) Kelly, Maulloo and Tan (1998) and MacKie Mason and Varian (1994). Our interest is in that part of the charge which reflects usage of the shared physical network resources of bandwidth and buffering, where it is the finiteness of these resources that is the principal binding constraint. In terms of the example above, such a constraint can be expressed (at least ....

MacKie-Mason, J. K. and Varian, H. R. (1994) Pricing the Internet. In B. Kahin and J. Keller, eds., Public Access to the Internet. Prentice-Hall, Englewood Cliffs, New Jersey.

....in a wholesale market that interconnects various separate grids. 17 The idea for a smart market originated with Vernon Smith and Charlie Plott [cite] Hal Varian and I have proposed a simple smart market for allocating spot delivery priorities on the Internet (MacKie Mason and Varian (1993) MacKie Mason and Varian (1994)) I have developed a smart market scheme for reserving resources in advance in a network running a connection oriented protocol like Asynchronous Transfer Mode (ATM) MacKie Mason (1994) 15 efficient allocations. The market is smart because the auctioneer clears prices not by simply ....

.... smart market for allocating spot delivery priorities on the Internet (MacKie Mason and Varian (1993) MacKie Mason and Varian (1994) I have developed a smart market scheme for reserving resources in advance in a network running a connection oriented protocol like Asynchronous Transfer Mode (ATM) (MacKie Mason (1994)) 15 efficient allocations. The market is smart because the auctioneer clears prices not by simply matching the highest bid to the lowest ask, but by repeatedly solving a complex programming problem that can handle multiple dimensions, varying quantities, and externalities. The smart ....

MacKie-Mason, J. K., and Varian, H. (1994). Pricing the Internet. In Kahin, B., and Keller, J. (Eds.), Public Access to the Internet. Prentice-Hall, Englewood Cliffs, New Jersey. Available from ftp://gopher.econ.lsa.umich.edu/pub/Papers.

....on their observed period t performance and their application requirements. This is a closed loop feedback system with at least a one period lag between the state of the network and the effect on the user inputs. 4. 3 Smart market pricing A smart market approach to adaptation has been proposed in [6]. A user sends packets to the network interface which include in each header a token indicating how much the user is willing to pay to get that packet onto the network in the current interval. Then, during the pricing interval, the network gateway sorts the bids on the incoming packets, and ....

....users pay just the congestion cost (the amount that the highest value denied packet would have paid for immediate transport) but they get to keep all of the excess value that they attribute to delivery above the cutoff bid. This form of pricing by auction has several nice properties, described in [6]. We call this mechanism a tight loop because the user sends willingness to pay along with the packet, and the network admission and pricing is determined based on those reports and the current state of network congestion, without creating a feedback delay. In practice, there might be a one period ....

[Article contains additional citation context not shown here]

J. MacKie-Mason and H. Varian, `Pricing the Internet,' in Public Access to the Internet, B. Kahin and J. Keller, eds., Prentice-Hall, Englewood Cliffs, NJ, 1995. Available from URL ftp://www.spp.umich.edu/ipps/papers/info-nets/Pricing Internet/Pricing the Internet.ps.Z

....a signal back to users based on the network utilization in period t. Users then decide how many cells to send in period t 1, based on their observed period t performance and their application requirements. 4.1. 3 Smart market pricing A smart market approach to adaptation has been proposed in [3]. A user sends cells to the network interface which include in each header an indication of how much the user is willing to pay to get that packet into the network in the current period. Then, during the pricing period, the network interface sorts the bids on the incoming packets, and admits to ....

....to the network. Thus, users pay just the congestion cost (the amount that the highest value denied packet would have paid for immediate transport) and they get to keep all of the excess value that they attribute to delivery. This form of pricing by auction has several nice properties, described in [3]. 4.2 Simulation details 4.2.1 Inelastic user This user has a delay bound on the traffic, but can tolerate only sending a fraction of the cells that are ready to go in the period in question. We assume that cells not sent in the period are useless to the user and are discarded. We assume that ....

J. MacKie-Mason and H. Varian, `Pricing the Internet,' in Public Access to the Internet, B. Kahin and J. Keller, eds., Prentice-Hall, Englewood Cliffs, NJ, 1995. Available from URL http://www.spp.umich.edu/spp/papers/jmm/ Pricing the Internet.pdf

....easy to use the computer to schedule some activities for off peak hours, leading to a shifting peaks problem. 9 In addition, so much traffic traverses long distances around the globe that time zone differences are important. Network statistics reveal very irregular time of day usage patterns (MacKie Mason and Varian (1994)) 8 See MacKie Mason and Varian (1993) 9 The single largest current use of network capacity is file transfer, much of which is distribution of files from central archives to distributed local archives. The timing for a large fraction of file transfer is likely to be flexible. Just as most fax ....

MacKie-Mason, J. K., and Varian, H. (1994). Pricing the internet. In Kahin, B., and Keller, J.

....responsible in part for the explosive increase of the Internet in recent years. Furthermore, it requires that tariffing, billing, and monitoring mechanisms be implemented in the Internet. Even though several have argued that the overhead associated with the billing mechanisms would be manageable [19], several problems might prove difficult to solve. For example, suggestions that tariffs should be based on traffic characteristics could be hard to implement in practice given the difficulties experienced in measuring and identifying such characteristics [18] One might then wonder whether the ....

J. MacKie-Mason, H. Varian, "Pricing the Internet", available from URL ftp://gopher.econ.lsa.umich.edu/pub/Papers/Pricing the Internet.ps.Z

....the outcome is essentially the same as the outcome of the standard English auction: the highest bidder gets the item but he pays (essentially) the second highest price. The Vickrey auction has been used in the computer science literature in Huberman and Hogg (1995) Rosenschein and Zlotkin (1994) MacKie Mason and Varian (1995), and no doubt in several other places. Since it is optimal for each person to reveal his or her true value, the Vickrey auction ensures that the item will be awarded to the individual with the highest willingness to pay. However, it does not maximize seller revenue: that problem is considerably ....

MacKie-Mason, J. K., and Varian, H. R. (1995). Pricing the internet. In Kahin, B., and Keller, J.

....of the Internet that constant returns to scale is not a bad approximation, at least for small changes in scale. Hence, the access prices we have described should serve as useful guides for capacity expansion. We have more fully described the role of congestion prices for guiding investment in MacKie Mason and Varian (1994), which includes a simple analytic model of congestion pricing and capacity expansion. Distributional aspects The issue of pricing the Internet is highly politicized. Since usage has been free for many years, there is a large constituency that is quite opposed to paying for it. One nice feature ....

MacKie-Mason, J. K., and Varian, H. (1994). Pricing the internet. In Kahin, B., and Keller, J.

No context found.

MacKie-Mason, J., and H. Varian "Pricing the Internet," in B. Kahin and J. Keller, editors, Public Access to the Internet, ACM, Boston, Massachusetts, May 1993.

No context found.

MacKie-Mason, J., and H. Varian "Pricing the Internet," in B. Kahin and J. Keller, editors, Public Access to the Internet, ACM, Boston, Massachusetts, May 1993.

No context found.

J. MacKie-Mason and H. Varian, "Pricing the internet," in Public Access to the Internet, B. Kahin and J. Keller, eds., pp. 269--314, MIT Press, 1995.

No context found.

MacKie-Mason, Jeffrey K., and Varian, Hal R.(1993).Pricing the Internet.

No context found.

J.K. Mackie-Mason and H.R. Varian, "Pricing the Internet", Proceedings of the Second International Conference on Telecommunication Systems Modeling and Analysis, pp. 378-393, Nashville, Tennessee, March 24-27, 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC