Orchestrating resources in 5G and beyond-5G systems will be substantially more complex than it used to be in previous generations of mobile networks. In order to take full advantage of the unprecedented possibilities for dynamic reconfiguration offered by network softwarization and virtualization technologies, operators have to embed intelligence in network resource orchestrators. We advocate that the automated, data-driven decisions taken by orchestrators must be guided by considerations on the cost that such decisions involve for the operator. We show that such a strategy can be implemented via a deep learning architecture that forecasts capacity rather than plain traffic, thanks to a novel loss function named α-OMC. We investigate the convergence properties of α-OMC, and provide preliminary results on the performance of the learning process in case studies with real-world mobile network traffic.
Industry and academia are the two major pillars that drive innovation and research, and historically, significant cross-pollination of ideas and expertise occurs between them. The goal of this panel is to discuss shared experiences, identify time-tested and new strategies, as well as role responsibilities in situations when industry and academia work together towards driving ambitious, open-ended research. The panel includes representatives from major research laboratories associated with the leading networking companies, startups, as well as industry consortiums for large wireless platform development projects committed to working with academic partners.
Cloud computing provides a kind of dynamic and scalable service on demand. However, clouds consume huge amountsof electrical energy. Meanwhile, delivering the negotiated QoS defined as Service Level Agreement (SLA) to users is necessary. A virtual machine placement strategy based on the equilibrium between energy and SLA is proposed. Aiming at dynamical changes of application workloads, an adaptive placement strategy RLWR based on robust local weight regression is presented, which decides the overload time of hosts dynamically according to the historical resource occupation of application workload. Then, two virtual machine migration selection algorithms, MPM and MNM are presented.The migrated virtual machines are deployed using bin-packing algorithm PBFDH. Contrasting to static algorithms such as STH, MPA and DVFS, virtual machines are not only deployed on fewer hosts in our algorithm, which promotes energy efficiency, but the load prediction can bring high-reliable QoS delivery and avoid overmuch SLA violations. Experimental results show that our strategy has an obvious effect on decreasing SLA violation under ensuring energy efficiency.
We consider an authentication process that makes use of biometric data or the output of a physical unclonable function (PUF), respectively, from an information theoretical point of view. We analyse different definitions of achievability for the authentication model. For the secrecy of the key generated for authentication, these definitions differ in their requirements. In the first work on PUF based authentication, weak secrecy has been used and the corresponding capacity regions have been characterized. The disadvantages of weak secrecy are well known. The ultimate performance criteria for the key are perfect secrecy together with uniform distribution of the key. We derive the corresponding capacity region. We show that, for perfect secrecy and uniform distribution of the key, we can achieve the same rates as for weak secrecy together with a weaker requirement on the distribution of the key. In the classical works on PUF based authentication, it is assumed that the source statistics are known perfectly. This requirement is rarely met in applications. That is why the model is generalized to a compound model, taking into account source uncertainty. We also derive the capacity region for the compound model requiring perfect secrecy. Additionally, we consider results for secure storage using a biometric or PUF source that follow directly from the results for authentication. We also generalize known results for this problem by weakening the assumption concerning the distribution of the data that shall be stored. This allows us to combine source compression and secure storage.
Network slicing to enable resource sharing among multiple tenants -network operators and/or services-is considered a key functionality for next generation mobile networks. This paper provides an analysis of a well-known model for resource sharing, the 'share-constrained proportional allocation' mechanism, to realize network slicing. This mechanism enables tenants to reap the performance benefits of sharing, while retaining the ability to customize their own users' allocation. This results in a network slicing game in which each tenant reacts to the user allocations of the other tenants so as to maximize its own utility. We show that, under appropriate conditions, the game associated with such strategic behavior converges to a Nash equilibrium. At the Nash equilibrium, a tenant always achieves the same, or better, performance than under a static partitioning of resources, hence providing the same level of protection as such static partitioning. We further analyze the efficiency and fairness of the resulting allocations, providing tight bounds for the price of anarchy and envy-freeness. Our analysis and extensive simulation results confirm that the mechanism provides a comprehensive practical solution to realize network slicing. Our theoretical results also fill a gap in the literature regarding the analysis of this resource allocation model under strategic players.
In mobile satellite network, scarcity of L/S frequency is an critical constraint for global communication. Moreover, the L/S frequency allocated for mobile satellite system (MSS) spans through more than 1GHz bandwidth with lower boundary 1610MHz and upper boundary 2690MHz. It entails several major technical challenges, such as high sampling rate required for wideband sensing, limited power and computing resources for processing load to seek for frequency through implementing wideband spectrum sensing. This paper investigates the issue of frequency availability in MSS, and wideband spectrum compressed signal detection approach is employed to obtain locations of active primary users (PUs) that should be avoided during frequency option. We propose novel discrete cosine transform based wideband spectrum compressed sensing scheme (DCT-WSCS) with significant improvement in detection probability and recovery accuracy compared with conventional discrete Fourier transform based wideband spectrum compressed sensing scheme (DFT-WSCS). Additionally, simulations are performed to demonstrate the performance of the proposed scheme in aspect of signal detection probability, reconstruction precision, and processing time.
Energy consumption has become one of the major challenges in the future communication networks. With the increased demand for capacity, the amount of energy needed by Radio Access Networks (RANs) grows significantly. Current research efforts propose solutions to increase network capacity at a much lower energy cost. For example in Distributed Base Station (DBS) or Cloud Radio Access Network (C-RAN), a set of Remote Radio Heads (RRHs) is connected to a baseband processing unit through an optical fiber which reduces the power needed to transmit the signal. Furthermore, thanks to centralized processing the number of sites decreases, thus network resources, including energy, can be better utilized through joint management. Another way to achieve this goal is deployment of low power nodes covering smaller areas than macro base stations, forming femtoor picocells. Even though small cells already reduce power consumption when compared to macro-sites, there is still room for improvement.
Wireless communication network consist nowadays of multiple standards, as well as cells of different sizes and coverage. Providing the best connection in such environment is a challenging task. We propose a new approach of solving the cell selection problem in heterogeneous networks. The method recursively applies efficient algorithms for bipartite graph matching to provide the final solution.
Network topology models have drawn tremendous interest from the research community. Traditionally, Internet modelling has been done at the AS or router level, in part because this information is most available from tomography and public databases such as Rocketfuel. We argue that physical topology analysis is important for a more complete understanding of Internet structure, particularly for insight into the resilience and survivability of infrastructure against attacks and natural disasters. However, complexity and lack of complete data sets has hindered accurate topology modelling. In this short paper, we show through a sample case that physical topologies have significantly different characteristics from traffic engineering and router-level overlays, and are important for analysis of geographically-correlated failures. Keywords-Internet topology, physical topologies, resilience, survivability
In this paper, we study two tightly coupled topics in online social networks (OSN): relationship classification and information propagation. The links in a social network often reflect social relationships among users. In this work, we first investigate identifying the relationships among social network users based on certain social network property and limited pre-known information. Social networks have been widely used for online marketing. A critical step is the propagation maximization by choosing a small set of seeds for marketing. Based on the social relationships learned in the first step, we show how to exploit these relationships to maximize the marketing efficacy. We evaluate our approach on large scale real-world data from Renren network, showing that the performances of our relationship classification and propagation maximization algorithm are pretty good in practice.
The penetration of smart pocket-size devices that provide constant Internet connectivity, such as mobile phones, has significantly changed the way people obtain, view and share information. Content provision is not anymore a prerogative to professionals; individuals are not solely customers, but also act as content generators and distributors. This shift in social behavior requires changes in the way information is delivered to target audiences in an efficient, interest-based and location-aware manner.This thesis explores a solution for opportunistic content distribution in a content-centric network that primarily targets content dissemination among mobile users in urban areas. The term ’opportunistic’ here refers to a concept which rejects the assumption of always-connected user devices and instead allows nodes to leverage sporadic contacts which occur when two neighbors come into direct radio communication range. Such communication mode allows data exchanges to occur in areas with little or no infrastructure; moreover, it is a potential solution for offloading the increasing traffic volumes observed by mobile operators.The contributions of this thesis lie in three areas. We first outline a general architecture and design for opportunistic content-centric networking. We implement our proposal on the Google Android platform, and provide application scenarios which illustrate the potential of mobile peer-to-peer communication. Our tests however show that energy consumption turns out to be a major issue for opportunistic networks. Therefore, our second effort is in the area of energy-efficiency. We propose a dual-radio architecture for opportunistic communication, and evaluate it through extensive simulations on realistic human mobility traces. Our final study lies in the area of content dissemination when nodes in the network act altruistically and are willing to solicit data on behalf of other participants. We propose a number of relaying and caching strategies, and evaluate them through simulations in environments that exhibit different churn levels.
In this paper, we propose a Tele-medicine application platform as a medical aid for patients suffering from Heart malfunction. We focus on heart diseases since they remain by far the major cause of death in the globe. Our solution utilizes the Satellite communication protocol DVB-RCS (Digital Video Broadcast- Return Channel Satellite), Wi-Fi, and the Network-on-Chip (NoC) technology. We utilize the 12-lead ECG biomedical technique to detect heart disorders via the biomedical NoC, which transmits the medical alarm and results via the biomedical network, ECG-BIONET. We do not investigate the DVB-RCS standard or Wi-Fi technology, but rather we try to utilize this technology, and we look at it from a performance point of view for our application by investigating three parameters, namely: delay, packet loss, and reliability. We follow a top down approach by looking at the needs of the application from a performance guarantee for our specific-purpose network.
