Wireless communication will play an increasingly important role in future factory automation and process control, where the presence of mobile autonomous devices is expected to grow. However, wireless links are prone to errors due to shadowing and multi-path fading, which is even more severe in dynamic environments. These problems can be attenuated by using a mesh backbone to which mobile node connect to, using a hand-off algorithm. This solution is particularly important under real-time requirements typically found in factory automation. In this paper, we devise the Double-Threshold Hand-off (DoTHa) algorithm, a novel hand-off mechanism that triggers a hand-off in various environmental conditions. As a case study, we carry out the tele-operation of a mobile robot through a wireless mesh network in an indoor setting, using a wireless chain network protocol (WICKPro-SRT) that supports soft real-time traffic. We empirically compared DoTHa with two existing hand-off algorithms based on single and double hysteresis margin. The results revealed that DoTHa achieves Data Delivery Ratio (DDR) close to 100% whereas the single hysteresis-based hand-off suffers from frequent disconnections, dropping DDR to 88%. The double hysteresis-based hand-off shows higher ping-pong effect than DoTHa, doubling the number of hand-offs in some scenarios.
In this paper, we show that in real-time switched Ethernet networks reducing the Maximum Transmission Unit (MTU) size may cause an increase or decrease in the response time of messages. This contradicting behavior arises an optimization problem for configuring the MTU size. We formulate the optimization problem in the context of the multi-hop HaRTES architecture, which is a hard real-time Ethernet protocol. As part of the solution, we propose a search-based algorithm to achieve optimum solutions. We modify the algorithm by presenting two techniques to reduce the search space. Then, we propose a heuristic algorithm with a pseudo-polynomial time complexity based on the search-based algorithm. We perform several experiments, and we show that the proposed heuristic results in an improvement regarding messages response times, compared with configuring the MTU to the maximum or minimum values. Moreover, we show in small network configurations that the heuristic performs as good as the search-based algorithm in many cases.
There is an increasing trend towards using switched Ethernet in real-time distributed systems due to features like absence of collisions and high throughput. Nevertheless, a few problems persist, in particular related to priority inversion and limited length in queues. In this paper we focus on a protocol which uses a master-slave technique over standard switched Ethernet in order to overcome such problems, namely FTT-SE protocol. We present an improved response time analysis for such a network and we compare, analytically and with simulations, the results achieved with Network Calculus on a worst-case scenario. We show that our proposed response time analysis gives tighter bounds compared to Network Calculus. Moreover, we compare the performance of different solutions to scale the FTT-SE protocol with respect to the bandwidth utilization. Finally, we propose a new architecture to improve the average performance of master-slave switched Ethernet networks.
Contemporary distributed embedded systems in many domains have become highly complex due to ever-increasing demand on advanced computer controlled functionality. The resource reservation techniques can be effective in lowering the software complexity, ensuring predictability and allowing flexibility during the development and execution of these systems. This paper proposes a novel end-to-end resource reservation model for distributed embedded systems. In order to support the development of predictable systems using the proposed model, the paper provides a method to design resource reservations and an end-to-end timing analysis. The reservation design can be subjected to different optimization criteria with respect to runtime footprint, overhead or performance. The paper also presents and evaluates a case study to show the usability of the proposed model, reservation design method and end-to-end timing analysis.
The resource reservation techniques provide effective means to lower the software complexity, ensure predictability and allow flexibility during the development and execution of complex distributed embedded systems. In this paper we propose a new end-to-end resource reservation model for distributed embedded systems. The model is comprehensive in such a way that it supports end-to-end resource reservations on distributed transactions with various activation patterns that are commonly used in industrial control systems. The model allows resource reservations on processors and real-time network protocols. We also present timing analysis for the distributed embedded systems that are developed using the proposed model. The timing analysis computes the end-to-end response times as well as delays such as data age and reaction delays. The presented analysis also supports real-time networks that can autonomously initiate transmissions. Such networks are not supported by the existing analyses. We also include a case study to show the usability of the model and end-to-end timing analysis with resource reservations.
Nowadays, switched Ethernet networks are used in complex systems that encompass tens to hundreds of nodes and thousands of signals. Such scenarios require multi-switch architectures where communications frequently occur in multiple hops. In this paper we investigate techniques to allow efficient multi-hop communication using HaRTES switches. These are modified Ethernet switches that provide real-time traffic scheduling, dynamic bandwidth management and temporal isolation between real-time and non-real-time traffic. This paper addresses the problem of forwarding traffic in HaRTES networks. Two methods have been recently proposed, namely Distributed Global Scheduling (DGS) that buffers traffic between switches, and Reduced Buffering Scheme (RBS), that uses immediate forwarding. In this paper, we discuss the design and implementation of RBS within HaRTES and we carry out an experimental validation with a prototype implementation. Then, we carry out a comparison between RBS and DGS using worst-case response time analysis and simulation. The comparison clearly establishes the superiority of RBS concerning end-to-end response times. In fact, with sample message sets, we achieved reductions in end-to-end delay that were as high as 80 %.
The number of mobile devices accessing the Internet is continuously increasing, creating new problems. For example, when many wireless clients try to access the same video stream in the same hot spot, multicast transmission must be used. However, multicast packets are unacknowledged and may present a high loss rate under adverse conditions which may jeopardize the quality of the streaming. In a severe scenario, packet losses may affect all video frames, preventing the clients from receiving the video stream at all. Therefore, we propose a novel reliable multicast protocol, tailored to streaming applications, that also scales to many clients under the same access point. It uses a feedback repeat request mechanism per client and such requests are adequately scheduled by the streaming server. This mechanism, which is the focus of this paper, also allows differentiating the quality of service provided to the clients, polling their feedback independently. We show experimental and simulation results that confirm the desired scalability and reliable streaming, with efficient bandwidth usage. The results also show a clear superiority in mitigating lost packets with respect to other approaches to the same problem, with benefits reaching and even going beyond one order of magnitude in packets recovery.
The increase in the number of mobile devices accessing the Internet creates new problems. A specific case is when a large number of wirelessclients try to access the same video stream in the same hot spot, which requires multicast transmission. However, multicast packets are unconfirmed and to compensate their higher loss rate we propose using a feedback repeat request mechanism per client. Then, the number ofclients is scaled up by adequately scheduling their feedback in the streaming server. This mechanism also allows managing the QoS provided to the clients controlling their feedback rate independently. Our preliminary results confirm the desired scalability with reliable streaming.
Ethernet is a promising candidate for networking real-time embedded systems such as automobiles, and it is already used in several embedded scopes, such as airplanes and trains. Moreover, its higher bandwidth enables applications made of dispersed embedded systems that exchange large amounts of data through the so-called Internet-of-Things, such as wide-area video sensing. In this realm, network reservations are an important design element that favor composability in the time domain thereby supporting the design of complex systems. However, reservations may impact negatively on the network bandwidth efficiency due to over-specification, particularly when they are designed to meet worst-case delay constraints. In this work, we use a specific Ethernet protocol that provides real-time reservations, namely FTT-SE, and we assess through extensive simulations the efficiency of a worst-case network delay analysis for sporadic reservations associated to asynchronous messages. Essentially, we compare the analytical worst-case delay with the one observed in the simulations using two data sets comprising up to 20000 and 100000 message sets, respectively, and we change the system configuration, namely properties of the message sets and network and protocol configuration parameters. We find that the analysis is accurate, i.e. matches observations, for a significant percentage of messages in the message sets (up to 60% on average). Conversely, we found that a small percentage of message sets (below 6%) generated rather pessimistic analytic estimates exceeding the observations by 6 times, thus with a negative impact on efficiency. The results we present in the paper, particularly their distributions, inform system designers of how to tune their designs to improve network bandwidth efficiency under strict timing constraints.
Complexity in the real-time embedded softwaredomain has been growing rapidly. The component-based softwaredevelopment approach facilitates the development process of suchsoftware systems by dividing a complex system into a numberof simpler components. Resource reservation techniques havebeen widely used for providing resources to real-time softwarecomponents. In this paper we target real-time components operatingon a distributed resource infrastructure. Furthermore,we target a class of software components which demonstratedynamic resource consumption behavior. A prime example ofsuch components is a multimedia software component. In thepaper, we present a framework supporting multi-resource endto-end resource reservations. We reserve resource bandwidths onboth processor resources as well as on the network resources. Theproposed framework utilizes a Multiple Input Multiple Output(MIMO) controller which adjusts the sizes of reservations trackingthe dynamic resource demands of the software components. Finally, we present a case study using a multimedia component todemonstrate the performance and efficiency of our framework.
—In this paper we present our ongoing work on developing a framework supporting adaptive resource reservations targeting component-based distributed real-time systems. The components may be spread over different resources in a distributed system. The proposed framework utilizes a reservationbased scheduling technique in which the sizes of reservations are adjusted during run-time to deal with dynamic resource demands of the software components. We present our modeling approach, we describe design options made and we present corresponding challenges.
Reliable multicast over wireless poses interesting challenges arising from the unreliable nature of the wireless medium. Recovering lost packets is particularly challenging in multicast scenarios since different receivers lose different packets. For this reason, simply retransmitting packets does not scale well with the number of receivers and particularly with the packet loss rate. A more efficient alternative is to use erasure codes to generate packets that can help many receivers at the same time. In this paper, we propose using online network coding to send coded packets that repair losses according to feedback reports sent by the clients. In particular, we propose using a recently developed scheduler for controlling feedback reports, which also allows differentiating the QoS provided to clients, and combine it with an online coding approach to provide novel stochastic guarantees of worst-case delay as required for QoS sensitive applications. We show preliminary simulation results that confirm the bounded decoding delay of our approach in a streaming application.
Small teams of cooperating robots have been shown to benefit from the increased reliability of synchronised message exchanges provided by TDMA-based schemes. However, such schemes may also impose a long propagation delay to communications between non neighbour robots. Such negative impact is further increased by long TDMA rounds, which favour reduced medium utilisation and energy consumption, and by decentralised mesh topologies, which increase the team coverage and layout flexibility. For small size teams, say up to 10 units, it is feasible to set-up a global TDMA framework as well as tracking the instantaneous network topology, making it available to each robot. In this work we use this knowledge to forward packets along their path within each TDMA slot. To that end, we present a novel communication protocol that combines global TDMA and multi-hop routing. It maintains the reliability benefits of the TDMA schemes while strongly reducing the end-to-end propagation delay of interactions between non neighbour robots. We validate our protocol with simulation results using OMNET++. In a worst-case topology scenario we achieved an end-to-end delay that can be as low as 35% that of a traditional TDMA implementation that forwards the packets to the immediate one-hop neighbours, only. In a concrete audio streaming application scenario reported in the literature for an alternative real-time token-passing protocol, our proposal achieves similar delays with significantly less management bandwidth.
Mobile devices in industrial settings became commonplace with different handheld equipment and mobile robots using COTS WiFi interfaces for communication. This growth in the use of the wireless medium increases the potential for overload and consequent long delays and high losses. Traffic segregation with different Quality of Service (QoS) classes can attenuate the problem but does not solve it within each class. In this paper we propose an extension to the Reconfigurable and Adaptive TDMA protocol to cope with generic traffic and improve the performance of highly loaded WiFi networks. We benefit from the self-synchronization feature of the protocol to create disjoint slots without clock synchronization which are then made available to the protocol stack through a TUN/TAP virtual interface. Then we propose a cooperative bandwidth management to adjust the traffic generated by each node through the virtual interface to match the available bandwidth, granting high bandwidth utilization while avoiding overload. We validate our approach with several practical experiments that confirm the capacity to measure and track the actual bandwidth available for communication through the protocol, removing the bandwidth consumed by interference.