Francesco Zampognaro

Analysis and performance evaluation of a burst-based TCP for satellite DVB RCS links

ESA Satlabs proposed a splitting architecture, named Interoperable-Performance Enhancing Proxy (I-PEP), which defines a protocol stack for the edges of a Digital Video Broadcasting-Return Channal over Satellite (DVB-RCS) link with the aim of improving performance of Transmission Control Protocol (TCP)-based applications. At the transport layer of I-PEPs, the Space Communications Protocols Standards-Transport Protocol (SCPS-TP) provides a reliable connection to upper layers, although resulting in very poor performance in tests. In fact, standard congestion-control mechanisms under-perform mainly due to long latency and Demand Assignment Multiple Access (DAMA) access schemes, especially in the case of short transfers, as for Web traffic, when optimum window may not be reached. In this paper, a burst-based TCP, named TCP Noordwijk (TCPN), is introduced to improve on these aspects. To evaluate performance, the protocol has been implemented on the Network Simulator NS-2. Definitively, details on the protocol design, implementation, and a vast gamut of results coming from simulations are reported.

Performance evaluation of TCP-based applications over DVB-RCS DAMA schemes

Transmission Control Protocol (TCP) performance over Digital Video Broadcasting-Return Channel via Satellite (DVB-RCS) standard is greatly affected by the total delay, which is mainly clue to two components, propagation delay and access delay. Both are significant because they are dependent oil the long propagation path of the satellite link. I-lie former is intrinsic and due to radio wave propagation over the satellite channel for both TCP packets and acknowledgements. It is regulated by the control loop that governs TCP. The latter is due to the control loop that governs the demand assignment Multiple access (DAMA) signalling exchange between satellite terminals and the network control center. necessary to manage return link resources. DAMA is adopted in DVB-RCS standard to achieve flexible and efficient use of the shared resources. Therefore, performance of TCP over DVB-RCS may degrade due to the exploitation of two nested control loops also depending oil both file selected DAMA algorithm and the traffic profile. This paper analyses the impact of basic DAMA implementation oil TCP-based applications over a DVB-RCS link for a large Set Of study Cases. To provide a detailed overview of TCP performance in DVB-RCS environment, the analysis includes both theoretical approach and simulation campaign. Copyright (C) 2009 John Wiley & Sons, Ltd.

Evaluation of TCP performance over emulated DVB-RCS scenario with multiple RCSTs

This paper analyses performance of competing TCP connections when running over different Return Channel Satellite Terminals (RCSTs) and competing for DVB-RCS return link bandwidth through DAMA mechanisms. A test campaign has been performed over a Linux-based satellite emulator supporting a multi-RCST configuration in a star-based architecture, with the main goal to evaluate TCP behavior in terms of both capacity utilization and fairness. In the proposed communication scenario, TCP performance relates on the joint effect of two nested control loops: TCP congestion/rate control loop and DAMA bandwidth assignment loop. Various TCP versions have been taken into account: Reno, Vegas, Westwood+, Hybla, Cubic and Noordwijk, a new TCP specifically designed for DVB-RCS point-to-point links. Rate Based Capacity Request (RBDC) has been selected as the access protocol.

Improving Performance of TCP-Based Applications over DVB-RCS Links

DVB-RCS links, as well as most of VSAT systems, to improve TCP performance usually consider a PEP solution. The SatLabs group of the European Space Agency issued I-PEP, selecting SCPS-TP as transport protocol, which supports two congestion control schemes: TCP Vegas (default) and van Jacobsons Slow Start and Congestion Avoidance. In both cases the most popular TCP-based applications, i.e. Web browsing and e-mail, achieve very poor performance especially when DAMA schemes run at the MAC layer. In this frame, TCP Noordwijk has been specifically designed to optimize performance of short TCP-based data transfers over DVB-RCS DAMA links, meeting the requirement to be compatible with I-PEP specifications. This paper presents an overview of TCP-based application criticalities over DVB-RCS links and shows how TCP Noordwijk improves performance.

Cross-Layer Architecture for a Satellite–Wi-Fi Efficient Handover

To achieve fully mobile communications, considering different environments and modern service requirements, a multiple-segment architecture is the most suitable to guarantee service continuity with acceptable performance. Handover (HO) procedures can be invoked either out of necessity (if the current network connection is going off) or to improve performance (if different bandwidth or quality of service is required). In this scenario, to provide uninterrupted communication services, efficient intersegment HO capability must be implemented. The architecture considered includes a satellite segment and a number of Wi-Fi hot spots. A mobile node (MN) can switch from a segment to other exploiting services of mobile Internet protocol (MIP). This architecture introduces great flexibility and ensures capillary coverage; it also strongly affects Transmission Control Protocol (TCP)-based application performance. To efficiently face HO consequences, particularly when the TCP runs as a transport protocol, an innovative protocol architecture based on cross-layer (CL) exchange of information is proposed. Analyses of TCP dynamics during HOs and the performance improvement introduced with the proposed CL architecture, evaluated through the network simulator Ns-2, are presented.

Joint DAMA-TCP protocol optimization through multiple cross layer interactions in DVB RCS scenario

Two aspects of DVB-RCS standard can worsen performance of TCP data connections: DAMA access scheme, since it introduces additional and variable delay to the already significant propagation delay and the adoption of Adaptive Coding on the return link to maximize bandwidth efficiency to face variable weather conditions, because it results in variable bandwidth allocation. Both aspects can severely impact TCP performance, especially for what concerns flow adaptation to varying channel conditions and channel usage efficiency. To optimize performance, in this paper cross-layer signaling among transport, MAC and physical layers of a DVB-RCS system is addressed. In particular MAC-TCP cross-layer is analyzed through the use of NS2 network simulator, showing the possible benefit in a DVB-RCS scenario.

Adaptable OPC-XML contracts taking into account network traffic

OPC-XML defines a standard method for accessing automation data in the industry field. This protocol relies on Web service communications allowing a great compatibility and a software/hardware independent definition. Nevertheless, this underlying protocol does not take into account temporal Quality of Service. Therefore, when a contract is defined between the client and the server, its execution has to be monitored to check if expected temporal properties are respected. In this paper, we propose to enrich the contract definition in order to define mandatory and optional quality of service. We propose an observation mechanism for the network congestion detection. Thanks to this information, it is possible to dynamically adapt the QoS to preserve the mandatory properties

TCP Noordwijk for High-Speed Trains

Mining abnormal patterns is important in many areas. With the prevalence of big data, in order to ensure efficiency, an algorithm named PPSpan (JOMP-based parallel Prefix Span) is proposed under the research of traditional serial sequential pattern mining methods. Firstly, redundant parameters are eliminated with grey correlation analysis. Secondly, outlier information is extracted according to the corresponding parameter threshold and each parameter is discretized with information entropy. Finally, PPSpan algorithm is employed to mine patterns. The algorithm can effectively mined the abnormal patterns from big dataset. Moreover, we verify the feasibility and effectiveness of the proposed method through an experimental analysis of a certain satellite data.Partial dynamic reconfiguration has become an important feature of FPGA-based systems as the number of applications which use this capability has increased. For systems using multiple partial bitstreams, the complexity of the target reconfigurable region, which often include heterogeneous blocks such as block RAMs and DSP blocks, makes it difficult to generate a unique bitstream which can be loaded into multiple locations in an FPGA. In this work we develop an FPGA architecture which allows for the placement of a partial FPGA design on the logic fabric even if the relative placement of heterogeneous blocks within the target region is not identical to the placement used to generate the bitstream for the partial design which include heterogeneous resources.Two main problems prevent a parallel garbage collection (GC) scheme with lock-based synchronization from providing a high level of scalability: the load imbalance and the runtime overhead of thread synchronization operations. These problems become even more serious as the number of available threads increases. We propose the Mark-Sharing algorithm to improve the performance of parallel GC using transactional memory (TM) systems. The Mark-Sharing algorithm guarantees that all threads access the shared resource by using both the task-stealing and task-releasing mechanisms appropriately. In addition, we introduce a selection manager that minimizes the contention and idle time of garbage collectors by maintaining task information. The proposed algorithm outperforms the prior pool-sharing algorithm of GC in the HTM, providing more than 90% performance improvement on average.In a scenario where people need to travel frequently for their business and leisure, spending a lot of time on high-speed trains, airplanes or ships, it is necessary to provide them a broadband Internet connection while moving. Broadband satellite systems, for their characteristics in terms of coverage and flexibility, can be the best solution for this type of scenarios. Nevertheless, although satellite systems allow a global coverage over the railways, the channel related to a high-speed train is characterized by frequent signal outages due to tunnels, steel covered bridges and powerline trellises. In particular, trellises, always and periodically present along the railroad, cause frequent obstructions of the line of sight, which implies packet loss that significantly affects TCP performance. In fact, standard TCP based on classical congestion control mechanisms is required to perform continuous restart of the algorithm. In this paper, the use of a new TCP protocol called TCP Noordwijk is proposed. It is based on burst transmission concept, which matches optimally with channel characteristics. The advantages with respect to the standard TCP will be shown.

Enhancing TCP Performance over Hybrid Wireless Terrestrial-Satellite Networks

We propose automatic synthesis of application specific instruction set processors (ASIPs). We use pipeline execution of multi-op machine-instructions, e.g., *(reg1*reg2) = (*reg3) + (*reg4) (C-syntax) an instruction with three memory stages and two arithmetic stages pipeline. The problem is, for a given set of loops, to find a pipeline configuration and a multi-op ISA that maximizes the IPC (instructions per cycle) while minimizing the resource usage and the cost of interconnections to the register-file of the resulting CPU. The algorithm is based on finding an efficient cover of a large graph by a small set of convex sub-graphs gis that are consistent with a given structure of a pipeline. Unlike previous works, gis are not synthesized to circuits that are executed in a co-processor mode but rather both gis and the rest of the program are executed by the same set of multiop pipeline units. In this way we eliminate the overhead associated with the co-processor mode of regular ASIPs but maintain high values of IPC of these ASIPs. The main advantage of using pipeline execution of multi-op versus VLIW instructions is shown to be the cost of interconnections between the CPUs execution units and the register file. Thus, we devise a grading function that for each possible multi-op pipeline configuration balance between the expected IPC (Instructions Per Cycle) and the complexity of the interconnections. Using this grading function we show that in most cases the VLIW configuration is not always the best choice.Data-intensive services have become one of the most challenging applications in cloud computing. The classical service composition problem will face new challenges as the services and correspondent data grow. A typical environment is the large scale scientific project AMS, which we are processing huge amount of data streams. In this paper, we will resolve service composition problem by considering the multi-objective data-intensive features. We propose to apply ant colony optimization algorithms and implemented them with simulated workflows in different scenarios. To evaluate the proposed algorithm, we compared it with a multi-objective genetic algorithm with respect to five performance metrics.This paper addresses the study of TCP performance over a heterogeneous network, where the access to IP based services is provided through the integration of terrestrial and satellite wireless technologies. This architecture is particularly suited for example for emergency communications, but presents some drawbacks in case TCP is used as transport protocol. The analysis of a solution for transport layer performance enhancement adopting TCP splitting agents at the edges of satellite link is discussed, using either a standard TCP or TCP Noordwijk for the inner connection between satellite terminal (ST) and satellite gateway (satGW). TCP-Noordwijk is a new transport protocol designed specifically for the satellite environment. This approach will be compared with a standard end-to-end TCP connection solution. Network Simulator NS-2 has supported the whole analysis and showed results which greatly demonstrate the advantages coming from the use of the split architecture and in particular adopting the TCP-Noordwijk transport protocol as inner connection.The continuous increase of the number of processing cores on die poses a new set of challenges to HPC applications programming including how to model, write, and verify software that has to use the full power of NoC-based manycore processors. Therefore, to simplify program development for the Single-chip Cloud Computer (SCC), it is desirable to have high-level, shared memory-based parallel programming abstractions (e.g., an OpenMP-like programming model). One of the key components of any similar programming model are barrier synchronization primitives, coordinating the work of parallel threads. To allow high-level barrier constructs to deliver good performance, we need an efficient implementation of the underlying synchronization algorithm. In this paper, we propose effective barrier synchronization implementations for shared-memory programming on non-cache-coherent cluster-on-chip represented by the Intel SCC. In particular, we present an extensive evaluation of the overhead associated with integrating barrier algorithms required for OpenMP runtime libraries on such a machine, validating several implementation variants that efficiently exploit the network topology and leveraging SCC-specific hardware. We provide a detailed evaluation of the performance achieved by different approaches by using micro-benchmarks.

A study on TCP error recovery interaction with Random Access satellite schemes

Random Access (RA) techniques have been included in the novel DVB-RCS2 standard due to an expected efficiency in addressing SCADA and M2M traffic. Such traffic relies on intermittent sources sending short messages on a periodic time scale. Efficiency is usually assessed in terms of layer-2 packet loss rate without taking into account indirect effects on upper layer protocols. In particular, TCP assumes packet losses as a congestion indication triggering actions aimed to reduce actual transmission rate. As a consequence, TCP over RA can be impaired by some cross-effects, which deteriorate the performance as experienced by the applications, in terms of Round Trip Time (RTT) and overall throughput. The scope of this paper is to analyze such dynamics, by applying a top-down approach, which includes TCP parameters tuning in the scenario configuration. To this aim, an NS2 module has been developed to simulate different flavors of random access techniques for GEO satellite channels integrated with sensor data sources, which produce intermittent messages and send them over TCP connections. Achieved results provide a first set of considerations on the best protocol stack configuration according to TCP traffic load.