Algorithmic Market Design: Spectrum Sales and BitTorrent Communities

Опубликовано 17 августа 2016, 1:38

Markets are a powerful tool for producing desirable outcomes in an engineered system. As computers allow us to build new and more complex markets, we need new algorithmic tools to understand and improve them. In this talk I'll discuss two examples of such markets. Until recently, if you wanted to start a business that requires using wireless spectrum, you would either have to pay tens of millions of dollars for an exclusive license or use one of the few congested frequencies that are available without a license (such as the one used by WiFi). Recently, people have begun investigating how auctions can be used to sell smaller amounts of spectrum. I'll present a new auction algorithm that takes advantage of the ability of devices to share spectrum to significantly improve both spectrum utilization and social welfare. Over the years, private file-sharing communities built on the BitTorrent protocol have developed their own policies and mechanisms for motivating members to share content and contribute resources. One popular mechanism is to require members to maintain a minimum ratio between uploads and downloads. By doing so, private communities effectively establish an economy. I'll present a year-long measurement study of DIMEΓÇöa community for sharing live concert recordingsΓÇöthrough which we find that users rationally react to cost differentials when making decisions on what files to consume. We observe significant disparities in the cost of new and old files, where by consuming newer files frequent visitors can quickly earn credit. We find that users compensate for the high cost of older files by downloading more copies of newer files, and by preferentially consuming older files during a ΓÇÿfree leechΓÇÖ period.