Cost-aware caching: Optimizing cache provisioning and object placement in ICN

Abstract

Caching is frequently used by Internet Service Providers as a viable technique to reduce the latency perceived by end users, while jointly offloading network traffic. While the cache hit-ratio is generally considered in the literature as the dominant performance metric for such type of systems, in this paper we argue that a critical missing piece has so far been neglected. Adopting a radically different perspective, in this paper we explicitly account for the cost of content retrieval, i.e. the cost associated to the external bandwidth needed by an ISP to retrieve the contents requested by its customers. Interestingly, we discover that classical cache provisioning techniques that maximize cache efficiency (i.e., the hit-ratio), lead to suboptimal solutions with higher overall cost. To show this mismatch, we propose two optimization models that either minimize the overall costs or maximize the hit-ratio, jointly providing cache sizing, object placement and path selection. We formulate a polynomialtime greedy algorithm to solve the two problems and analytically prove its optimality. We provide numerical results and show that significant cost savings are attainable via a cost-aware design.

Coupling caching and forwarding

“…From a technical viewpoint, it follows that future ICN literature should not limitedly consider a naïve SPR,LCE,LRU strategy, as it offers a too weak candidate for comparison, but also consider iNRR,LCD,LRU or even better alternatives [37]. Comparison with optimal strategies [3] is still missing: gauging the distance from optimum would help in understanding the extent of gains that are still possible beyond iNRR,LCD,LRU . From an economic viewpoint, business considerations will answer whether such gains are economically worth the deployment of ICN -yet, business considerations should be taken on the ground of all relevant technical information.…”

Design and evaluation of cost-aware information centric routers

“…In an ongoing related effort [5], we show these to be contrasting objectives in an optimization framework. In this work, we instead focus on a complementary perspective: the design of a distributed cost-aware mechanism, whose performance approaches the one gathered by the solution of a centralized optimization problem.…”

“…We finally find the minimal ISP cost by solving the optimization problem formalized in [5], where we minimize the cost incurred by an ISP by storing in the cache, a priori, objects o with the largest product of cost times popularity λ oπo.Since we use the optimum as a reference against our design, we deem its full formulation to be outside the scope of the paper, and refer the interested reader to [5] for more details.…”

External Assessment of QoS Provisioning in Distributed Cloud Services

“…The latency in delivery of a content to the clients: say, < 5 , is a QoS parameter prescribed by clients. A service-layer algorithm places proxy nodes in the distribution tree to store various contents, whereupon a 'push' or 'pull' strategy is employed to deliver content to the clients [3], [4]. To meet the latency constraints and security needs, one or more nodes in the distribution topology function as proxies for maintaining a local copy of , so that clients pull from a proxy node in their geographic vicinity.…”