In this position paper we introduce Virtual Telco, a comprehensive effort to simplify the management of deployed telecommunication services by using a Cloud computing approach. The main objective of Virtual Telco is to replace the costly dedicated hardware implementing several centralized control plane functions and other services with distributed solutions that may be allocated on-demand over a pool of dependable, dynamically contracted computing and networking resources that are easy to manage. Virtual Telco relies on state-of-the-art techniques in the domains of virtualization and distributed systems to meet the challenging criteria of reliability, scalability, and timeliness required by present and future telecommunication standards and services. As a representative example of a Virtual Telco application, we propose the case of the distributed mobility management entity (MME) for next-generation LTE cellular networks.

In distributed storage systems, erasure codes represent an attractive solution to add redundancy to stored data while limiting the storage overhead. They are able to provide the same reliability as replication requiring much less storage space. Erasure coding breaks the data into pieces that are encoded and then stored on different nodes. However, when storage nodes permanently abandon the system, new redundant pieces must be created. For erasure codes, generating a new piece requires the transmission of k pieces over the network, resulting in a k times higher reconstruction traffic as compared to replication. Dimakis proposed a new class of codes, called regenerating codes, which are able to provide both the storage efficiency of erasure codes and the communication efficiency of replication. However, Dimakis gave only a theoretical description of the codes without discussing implementation issues or computational costs. We have done a real implementation of random linear regenerating codes that allows us to measure their computational cost, which can be significant if the parameters are not chosen properly. However, we also find that there exist parameter values that result in a significant reduction of the communication overhead at the expense of a small increase in storage cost and computation, which makes these codes very attractive for distributed storage systems.

Redundancy is the basic technique to provide reliability in storage systems consisting of multiple components. A redundancy scheme defines how the redundant data are produced and maintained. The simplest redundancy scheme is replication, which however suffers from storage inefficiency. Another approach is erasure coding, which provides the same level of reliability as replication using a significantly smaller amount of storage. When redundant data are lost, they need to be replaced. While replacing replicated data consists in a simple copy, it becomes a complex operation with erasure codes: new data are produced performing a coding over some other available data. The amount of data to be read and coded is d times larger than the amount of data produced. This implies that coding has a larger computational and I/O cost, which, for distributed storage systems, translates into increased network traffic. Participants of peer-to-peer systems have ample storage and CPU power, but their network bandwidth may be limited. For these reasons existing coding techniques are not suitable for P2P storage. This work explores the design space between replication and the existing erasure codes. We propose and evaluate a new class of erasure codes, called hierarchical codes, which aims at finding a flexible trade-off that allows the reduction of the network traffic due to maintenance without losing the benefits given by traditional codes.

Distributed storage systems provide data availability by means of redundancy. To assure a given level of availability in case of node failures, new redundant fragments need to be introduced.

This paper presents a generalized feature projection scheme which allows each feature dimension to be classified in a set of 1 to M classes, where M is the total number of classes. Our method is an extension of the classical full-space null-space approach where each dimension can only be classified in either M classes or 1 class. We believe that this more general formulation allows for a better trade-off of number of parameters versus model complexity, which in turn should provide better classification. We first tested GLDA on TIMIT and obtained an improvement up to 1% in phone classification rate over the best HLDA classifier. Preliminary results on Wall Street Journal 20K also show an improvement over the best HLDA system of about 0.2% absolute.