Market-based cloud resource allocation

From Lsdf
Revision as of 14:51, 10 March 2016 by Diana.gudu (talk | contribs) (Created page with "Back to the list of topics = Description = In this research project at SCC, we are working towards a ''Scalable resource allocation in large...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Back to the list of topics


In this research project at SCC, we are working towards a Scalable resource allocation in large-scale, dynamic environments by using a decentralised, market driven approach.

As cloud computing becomes an established technology, there are new challenges [0][1] related to the multitude and diversity of cloud providers (in terms of the type of resources they offer, the prices, the quality of service etc). This makes it difficult for customers to select the best cloud provider that fits their requirements for the best price. This is where so-called "brokers" come into play to automate the provider selection and deployment.

Our research aims to:

  1. make such brokering more scalable through decentralisation (using multi-agent systems)
  2. introduce market mechanisms such as auctions for more flexibility (a simple example already in used in production is Amazon Spot instance [2], where virtual machines are auctioned off at fluctuating prices based on supply and demand).

Your tasks would be to investigate and implement different auction algorithms for resource allocation.


  • implement different combinatorial auction algorithms
  • evaluate algorithms' performance and quality under different scenarios
  • identify potential for parallelization
  • integrate thesse algorithms into our multi-agent system simulator (based on RepastHPC [3])
  • evaluate scalability by running simulations


  • good knowledge of C++
  • Python/Bash or other scripting language
  • some theoretical computer science background (e.g. computational complexity theory) would be a plus


[0] R. Buyya, C. S. Yeo, and S. Venugopal. Market-oriented cloud computing: Vision, hype, and reality for delivering it services as computing utilities. HPCC 2008. doi:10.1109/HPCC.2008.172
[1] M. Armbrust, A. Fox, R. Griffith, et al. Above the Clouds: A Berkeley View of Cloud Computing. Technical report, University of California at Berkeley, February 2009.
[2] Amazon Spot instances
[3] RepastHPC
[4] S. De Vries and R. V. Vohra. Combinatorial auctions: A survey. INFORMS Journal on computing, 2003. doi:10.1287/ijoc.