Wormhole Networks Research Articles

A deadlock detection mechanism for true fully adaptive routing in regular wormhole networks

Different from traditional deadlock avoidance schemes, deadlock detection and recovery-based routing algorithms in wormhole networks have gained attention due to low hardware complexity and high routing adaptability. By its nature, current deadlock detection techniques based on time-out accompany unignorable number of false deadlock detections especially in a heavily loaded network or with long packet size and may mark more than one packet in a deadlock as deadlocked. This would saturate the resources allocated for recovery, making deadlock recovery schemes less viable. This paper proposes a simple but more accurate deadlock detection scheme which is less dependent on the time-out value. The proposed scheme uses a control packet to find a potential cyclic dependency between packets and presumes a deadlock only upon finding such dependency. The suggested scheme considerably reduces the probability of false deadlock detections over previous schemes, thus enabling more efficient deadlock recovery and higher network throughput. Simulation results are provided to demonstrate the efficiency of the proposed scheme.

Consensus with Dual Fallible Wormhole Network

Consensus with Dual Fallible Wormhole Network

A family of mechanisms for congestion control in wormhole networks

Multiprocessor interconnection networks may reach congestion with high traffic loads, which prevents reaching the wished performance. Unfortunately, many of the mechanisms proposed in the literature for congestion control either suffer from a lack of robustness, being unable to work properly with different traffic patterns or message lengths, or detect congestion relying on global information that wastes some network bandwidth. This paper presents a family of mechanisms to avoid network congestion in wormhole networks. All of them need only local information, applying message throttling when it is required. The proposed mechanisms use different strategies to detect network congestion and also apply different corrective actions. The mechanisms are evaluated and compared for several network loads and topologies, noticeably improving network performance with high loads but without penalizing network behavior for low and medium traffic rates, where no congestion control is required.

A New Methodology for Modelling of Sand Wormholes in Field Scale Thermal Simulation

Abstract Sand wormholes created by the concurrent production of sand and heavy oil are instrumental in improving oil rates and ultimate recovery. The wormholes form high permeability conduits for flow of oil and more sand. The wormholes may extend several hundred metres away from each well and connect with other wells' wormholes. Efforts have been made to predict the wormhole network based on growth of each wormhole's tip. In this paper, we use a fractal geostatistical method to pregenerate the wormhole network. This wormhole network is then incorporated into a full field thermal simulation model. A dual porosity-dual permeability grid is used, one for the reservoir pay and one for the wormhole network. The wormhole grid cells have the same pore volume as the wormholes and a sparse network structure. With this wormhole representation in the simulation model, we were able to match thermal responses of cold production wells from a steam injection pilot due to the wormholes conducting hot fluid from the steam injectors. As a result, this modelling concept for wormholes may be used to investigate post cold production thermal recovery methods as well as for thinner reservoirs not suitable for Steam Assisted Gravity Drainage (SAGD) where the wormholes may act as natural horizontal wells. Introduction Sand production occurs with heavy oil production in unconsolidated sand formations. If sand production is stopped with screens or filters, this often results in near total loss of production from the well. With the use of progressive cavity pumps, sand production can be encouraged, resulting in sand cuts that can be as high as 30 – 40% initially before dropping to about 5%. The production of sand results in open holes, whimsically called wormholes, that stretch far into the formation away from the well. The productivity of the well rises from the average 4 to 5 m3/d to as high as 15 to 20 m3/d as the wormholes form high permeability conduits for flow of oil and more sand. This production process is called Cold Heavy Oil Production with Sand or CHOPS. Many papers have been published on the mechanics of sand wormhole development. Recently, Y. Wang and C. Chen(1) describe a reservoir model coupled with a detailed wormhole propagation model. The model also incorporates slurry transport with three phase fluid flow. This appears to be a single well model. Research work at the Alberta Research Council on modelling cold production was published by Sawatzky et al.(2). They provide a review of cold production technology as well as their modelling efforts in combining a comprehensive sand production with a fluid flow simulator. It would appear that this too was a single well model predicting the development of wormholes from a well using criterion for sand failure at the tip of a wormhole. These two papers also list extensive bibliographies which will not be repeated here. We are, however, more concerned with the effect of a developed network of wormholes on recovery from a reservoir than with trying to predict the growth of each wormhole based on stresses at its tip.

Modelling of Sand Transport Through Wormholes

Abstract One of the key mechanisms in the cold production (CHOPS) process is the development of high permeability channels (wormholes), which increase the access to the reservoir. In order to model the growth of wormholes in the field, it is necessary to predict the flow of sand along the wormholes. The laminar flow of sand and oil through an open channel within a wormhole was modelled using an analytical Bingham Mohr-Coulomb model. This model was tested by comparing its predictions to oil and sand flow rate measurements obtained in pipe flow experiments reported in the literature. The model for sand transport was used in developing a field wormhole growth model. Introduction In Western Canada, well operators have observed that encouraging sand production along with the oil, in primary recovery, can lead to the economical production of heavy oil from thin reservoirs. Field tests and laboratory experiments suggest that the enhanced recovery associated with massive sand production can best be explained by the development of high permeability channels (wormholes), which access the reservoir(1). The existing cold production models can be essentially divided into three groups.Equivalent Permeability Models: In these models, the production of sand was thought to lead to the development of a radial higher permeability (dilated) zone which would start at the vertical well and grow into the formation(2–5). The earlier models assumed that the permeability of the dilated zone was constant(2). Denbina et al.(3) concluded that both a dynamically enhanced absolute permeability of the formation with sand production and a suppressed gas relative permeability were required to history match the oil production from typical cold production wells. Kumar and Pooladi-Darvish(4) went further in this approach by also history matching the sand production from a typical cold production well. Wang and Chen(5) developed a cold production model which also takes into account the sand transport within wormholes. They history matched the cumulative oil and sand for a typical cold production well.Wormhole Network (Fractal) Models: In this category of models, the enhanced permeability zone caused by sand production is assumed to be caused by a fractal network of wormholes(6–9). The number and density of the wormhole network was assumed to be large enough such that on a larger scale, the region from which sand is produced can be assumed to be uniform. This required that the wormholes be small and that significant branching occur. Reasonable history matches of the oil and sand production rates were obtained for a set of cold production wells(8). Tan et al.(6) also matched the temperature in producing wells during steam injection. The wormhole network was modelled using the Diffusion Limited Aggregation (DLA) fractal geostatistical approach. A semi-analytical cold production model based on a fractal distribution of wormholes was also developed by Liu and Zhao(9).Linear Channel Models: Loughead and Saltuklaroglu(10) interpreted the pressure build-up tests in cold production wells as indicative of linear flow through a high permeability channel.

An Integrated Experimental Study of Foamy Oil Flow During Solution Gas Drive

Abstract One of the important mechanisms in the primary production of heavy oil reservoirs under solution gas drive is the foamy oil flow effect. This form of flow occurs only under certain combinations of capillary, viscous, and gravity forces. Several studies have been carried out to examine this effect on the recovery process, but it remains poorly understood and difficult to model. The objective of this study was to examine the roles of capillary, viscous, and gravity forces in foamy oil flow. Primary depletion experiments were conducted in a 2 m long sand pack holder to measure the oil and gas production at different depletion rates. The effect of gravitational forces was investigated by comparing the experiments conducted in horizontal and vertical orientations of the 2 m long sand pack. In the vertical orientation, the solution gas drive performance was evaluated with production from the bottom end of the sand pack, as well as with production from the top end. The experimental results are presented in terms of the oil and gas production behaviour affected by the capillary number and gravity forces. The results indicate that the gravity forces do not significantly influence the ultimate oil recovery when the oil is produced under foamy oil flow. Even at the lowest depletion rate used in these tests, when the oil flow and gravity are in the same direction, gravity segregation of gas did not improve the oil recovery significantly. Recovery efficiency and the critical gas saturation at which free gas flow starts are primarily dependent on the capillary number. Introduction Some of the heavy oil reservoirs in the world show anomalous primary production under solution gas drive. Smith(1) was one of the first researchers who presented a detailed analysis of such unusual production behaviour. Since then, it has been the subject of many studies, due to its importance in making reliable projections of future production and optimization of field operations. These studies can be divided into three categories. The first category focuses mainly on the geomechanical aspects(2, 3). It is believed that the production of sand with oil improves the inflow performance by creating wormholes (high permeability channels). Experimental and theoretical studies(3, 4) have shown that the high pressure gradients present during cold production can overcome the cohesive strength of the sand matrix and this leads to the creation of wormholes. However, the mechanics of wormhole network development are not yet clear. The second category focuses on multiphase flow behaviour, and can be further divided into two subcategories. First, there are the conventional two-phase flow models(5, 6), which suggest that solution gas drive, whether in light oils or in heavy oils, is describable by the critical gas saturation and oil-gas relative permeability. In these, the high oil recovery is attributed mainly to low gas mobility. However, a good history match of the field performance is not obtained by using low gas mobility; it also requires enhanced permeability far from the wellbore due to sand production(7).

Analysis of true fully adaptive routing with software-based deadlock recovery

Several analytical models of fully adaptive routing (AR) in wormhole-routed networks have recently been reported in the literature. All these models, however, have been discussed for routing algorithms with deadlock avoidance. Recent studies have revealed that deadlocks are quite rare in the network, especially when enough routing freedom is provided. Thus the hardware resources, e.g. virtual channels, dedicated for deadlock avoidance are not utilised most of the time. This consideration has motivated researchers to introduce fully adaptive routing algorithms with deadlock recovery. This paper describes a new analytical model of a true fully AR algorithm with software-based deadlock recovery, proposed in Martinez et al. (Software-based deadlock recovery techniques for true fully AR in wormhole networks, Proc. Int. Conf. Parallel Processing (ICPP’97), 1997, p. 182) for k-ary n-cubes. The proposed model uses the results from queueing systems with impatient customers to capture the effects of the timeout mechanism used in this routing algorithm for deadlock detection. Results obtained through simulation experiments confirm that the model predicts message latency with a good degree of accuracy under different working conditions.

Fc3d: flow control-based distributed deadlock detection mechanism for true fully adaptive routing in wormhole networks

Two general approaches have been proposed for deadlock handling in wormhole networks. Traditionally, deadlock-avoidance strategies have been used. In this case, either routing is restricted so that there are no cyclic dependencies between channels or cyclic dependencies between channels are allowed provided that there are some escape paths to avoid deadlock. More recently, deadlock recovery strategies have begun to gain acceptance. These strategies allow the use of unrestricted fully adaptive routing, usually outperforming deadlock avoidance techniques. However, they require a deadlock detection mechanism and a deadlock recovery mechanism that is able to recover from deadlocks faster than they occur. In particular, progressive deadlock recovery techniques are very attractive because they allocate a few dedicated resources to quickly deliver deadlocked messages, instead of killing them. Unfortunately, distributed deadlock detection is usually based on crude time-outs, which detect many false deadlocks. As a consequence, messages detected as deadlocked may saturate the bandwidth offered by recovery resources, thus degrading performance. Additionally, the threshold required by the detection mechanism (the time-out) strongly depends on network load, which is not known in advance at the design stage. This limits the applicability of deadlock recovery on actual networks. We propose a novel distributed deadlock detection mechanism that uses only local information, detects all the deadlocks, considerably reduces the probability of false deadlock detection over previously proposed techniques, and is not significantly affected by variations in message length and/or message destination distribution.

A foundation for designing deadlock-free routing algorithms in wormhole networks

This article provides necessary and sufficient conditions for deadlock-free unicast and multicast routing with the path-based routing model in interconnection networks that use the wormhole switching technique. The theory is developed around three central concepts: channel waiting, False Resource Cycles , and valid destination sets . The first two concepts are suitable extensions to those developed for unicast routing by two authors of this article; the third concept has been developed by Lin and Ni. The necessary and sufficient conditions relax the requirements for deadlock-free routing, compared to techniques that provide only a sufficient condition. These necessary and sufficient conditions can be applied in a straightforward manner to prove deadlock freedom of newly developed adaptive routing algorithms for collective communication, which in turn will help in developing efficient and correct routing algorithms. The latter point is illustrated by developing two routing algorithms for multicast communication on 2D mesh architectures. The first algorithm uses fewer resources (channels) than an algorithm proposed in the literature but achieves the same adaptiveness. The second algorithm provides fully adaptive routing for both unicast and multicast messages.

Real-time wormhole channels

For real-time communication, we must be able to guarantee timely delivery of messages. In recent years, improvements in technology have made possible switch-based local area networks (LANs) and system area networks (SANs) that use wormhole switching, a pipelined switching technique which permits significantly shorter network latencies and higher throughputs than traditional store-and-forward packet switching. This paper proposes a model for real-time communication in such wormhole networks based on the use of real-time wormhole channels, which are simplex virtual circuits in wormhole networks with certain real-time guarantees. A distinguishing feature of our model is that it can be used in existing wormhole networks without any special hardware support. Preliminary delay analysis and properties are shown for the proposed real-time wormhole channel model. A practical quadratic time complexity algorithm is shown for determining the feasibility of a set of wormhole channels. Finally, as an example of the utility of our model, actual parameter values obtained from experiments on a Myrinet switch-based network are used to determine if real-time guarantees are possible for an example set of real-time traffic streams intermixed with nonreal-time traffic.

Preliminary Results From a Solvent Gas Injection Field Test in a Depleted Heavy Oil Reservoir

Abstract A field test was run during 1997 and 1998 to collect preliminary data on a solvent gas injection process. The site selected for the test was a "typical" Frog Lake, Alberta, Cummings formation reservoir that had been depleted using PC pump-based cold production methods on both 4 hectare (10 acre) and 8.1 hectare (20 acre) spacing. At the time of solvent injection, area average recovery from the test location was 9.5%, and the existence of wormholes in the reservoir was strongly suggested by regional well-to-well water migration and production of reservoir sand at high initial rates and cumulative volumes. A solvent containing 33 volume % propane and 67 volume % methane was injected at two converted central producers until cumulative volumes of 2 million m3 and 4 million m3 were achieved. This paper discusses field observations during the injection, soak, and production periods. A number of circumstances contributed to a poor economic result, but the operational and technical information obtained should prove helpful to operators considering application of solvent injection processes. Introduction Cold Production in Western Canadian Regional Heavy Oil Sands PC pump-based cold production has expanded rapidly following initial development in the 1980s. This exploitation methodology provides a large percentage of the produced oil volume for most Western Canadian heavy oil producers. Some producers use only this method. Despite continued efforts to improve PC pump-based cold production technology, current methods generally leave 80 to 95% of the OOIP behind at economic limit. While this is a large oil-inplace target for follow-up EOR processes, the cold production process appears to have strongly altered reservoir conditions. Wormhole channels have been implied and described on the basis of observed high produced sand volumes and rapid migration of edge water and injected fluids and tracers(1–3). The reservoirs are generally pressure depleted. Solution gas, which appears to help power production through mechanisms described as either foamy oil flow(4–6) or more recently as extremely low gas mobility(7), often appears to "blow down" at the end of a well's life. Water influx, likely through wormhole networks, sometimes occurs at wells distant from the original oil/water contact. Thermal Production in Western Canadian Regional Heavy Oil Sands During the 1960s, operators attempted steam processes in selected heavy oil regional sands to recover a higher fraction of OOIP. It was soon determined that the typical H-40 casings with non-thermal cement would not withstand the resulting thermal stresses. Insulated tubing strings with or without packers were tried, but these strings were expensive, fragile, and often did not achieve theoretical performance in the field. Thermal completions in newer wells avoided wellbore failure due to the use of premium tubulars and collars, but other problems occurred. Often the steam had to be injected at pressures exceeding the formation parting (fracture) pressure to achieve acceptable heat transfer rates to the reservoir. This resulted in rapid steam channeling to neighboring wells, or to neighboring formations.

A Model for Sand Transport Through a Partially Filled Wormhole in Cold Production

Abstract We propose a simple model of flow in a partially filled wormhole, where layers of oil, slurry and immobile sand can coexist. The slurry at the bottom of the wormhole is assumed to behave as a Bingham material, i.e., it yields when the shear stress exceeds the yield stress. The yield stress is assumed to increase; with depth due to the weight of the overlying sand. A Mohr-Coulomb equation is used to calculate the yield stress within the slurry and immobile sand layers. Linearized Navier- Stokes equations are solved to calculate the oil and mobile sand flow rates. Different characteristic oil and sand flow patterns are studied. Oil and sand flow rates through the wormhole are calculated as functions of pressure gradient and rheological parameters. This simple model could be used to estimate sand transport in a partially filled wormhole, complementing a sand transport model for uniformly filled wormholes(4). These wormhole flow models can be incorporated into a field scale model for cold production. Introduction Cold production is a non-thermal primary heavy oil recovery process in which sand production is encouraged. Previous field and laboratory studies(1–3) suggest that high permeability channels (wormholes) develop within the formation starting at the perforations in a cold production well. These wormholes provide high effective fluid mobility, leading to a higher oil recovery as comparedto conventional primary recovery without sand production. According to previous models we developed(4, 5), in the first few months of cold production, wormholes grow rapidly, forming a wormhole network. The produced sand cuts are high during this period. Based on laboratory experiments, during this early stage, most wormholes are uniformly filled with slurry, a mixture of oil and liquefied sand. At later stages of cold production, however, the expansion of the wormhole network zone slows down, resulting in reduced sand cuts. In previous experiments conducted at ARC, we observed the creation and extension of a sand-free zone in the upper part of a wormhole after the wormholes stopped growing(3). The flow behaviour under this circumstance is very different from that in a uniformly filled wormhole. In this work, we use a simplified description of the layered flow in partially filled wormhole. A comparison with experimental results indicates that this simple model could be used for the estimation of flow properties in a partially filled wormhole, complementing the sand transport model for a fully filled wormhole and the wormhole network model. Backgroound The actual geometry of a wormhole is complicated. Therefore, we assumed a plane geometry, as is shown in Figure 1. In the region 0 ≤y ≤h, we assume that sand and oil flow with at the same velocity; Equation (Available In Full Paper) The rheology of the slurry is described by the Bingham Equation(8): Equation (Available In Full Paper) where, sign(π) = 1 for positive π, and -1 for negative π. The righthand side of this equation is zero if the magnitude of the shear stress is less than the yield stress of the slurry.

A new injection limitation mechanism for wormhole networks

Wormhole switching has been widely applied to the interconnection networks of parallel systems as well as System Area Networks, and Local Area Networks, largely because of its efficiency and performance merits. Examples include the Myrinet of Myricom Inc as well as most of the newly developed parallel systems. True Fully Adaptive Routing (TFAR) Algorithms have demonstrated their suitability for wormhole switched networks due to their unrestricted Adaptivity and moderate resource requirements. Wormhole switching has proven to be the most popular switching technique targeted for interconnection networks of message-passing multicomputers as well as SANs, and LANs. TFAR Algorithms have also been gaining favor for application in wormhole switched networks due to their highly adaptive and moderate hardware requirements. Wormhole switched networks have associated drawbacks however, as they generally suffer from performance degradation beyond the saturation point due to channel congestion. Fully adaptive algorithms are vulnerable to cyclic dependencies, which are precursors to deadlock formations. Consequently the frequent occurrence of deadlocks can further degrade the performance and stability characteristics of these networks. Injection limitation techniques were recently introduced in an attempt to countermeasure these drawbacks and effectively contain their impact on the performance of the network. This paper proposes a new injection limitation mechanism and its performance evaluation. The new mechanism is named Congestion Level Injection Control (CLIC). This mechanism attempts to provide a solution for these problems and improve the overall performance of the network. The new mechanism is centered on congestion level estimation in the network using only local information at each node. The mechanism subsequently prevents the injection of new packets if the network is deemed to be highly congested or possibly close to its saturation point. The performance of the CLIC mechanism has been compared with other competing schemes. Our results have shown that CLIC has superior performance when compared to other competing schemes.

A cost-effective approach to deadlock handling in wormhole networks

Wormhole networks have traditionally used deadlock avoidance strategies. More recently, deadlock recovery strategies have begun to gain acceptance. In particular, progressive deadlock recovery techniques allocate a few dedicated resources to quickly deliver deadlocked packets. Deadlock recovery is based on the assumption that deadlocks are rare; otherwise, recovery techniques are not efficient. Measurements of deadlock occurrence frequency show that deadlocks are highly unlikely when enough routing freedom is provided. However, networks are more prone to deadlocks when the network is close to or beyond saturation, causing some network performance degradation. Similar performance degradation behavior at saturation was also observed in networks using deadlock avoidance strategies. In this paper, we take a different approach to handling deadlocks and performance degradation. We propose the use of an injection limitation mechanism that prevents performance degradation near the saturation point and, at the same time, reduces the probability of deadlock to negligible values. We also propose an improved deadlock detection mechanism that uses only local information, detects all deadlocks, and considerably reduces the probability of false deadlock detection over previous proposals. In the rare case when impending deadlock is detected, our proposal consists of using a simple recovery technique that absorbs the deadlocked message at the current node and later reinjects it for continued routing toward its destination. Performance evaluation results show that our new approach to handling deadlock is more efficient than previously proposed techniques.

Supporting TCP connections in wormhole routing and ATM networks

An interesting lightweight switching technique for local area and campus environments is wormhole routing, in which the head of a packet, upon arriving at an intermediate switch, is immediately forwarded to the next switch on the path. The current proposals and products of wormhole routing networks provide a connectionless and lossless transport of variable-size data units, aiming at minimal latencies, but without any quality-of-service guarantee. Wormhole networks therefore push at an extreme the networking paradigm of the Internet, adopting technical solutions for high-speed networking that are opposite to those taken in ATM. The aim of this paper is to perform a simulation-based comparison between the wormhole routing and ATM transport techniques in high-speed networking scenarios. The behavior of the two information transport techniques is studied in mesh topologies where a few interfering TCP connections in overload interact with a uniform background traffic. Our results show that a larger hardware and software complexity is required to ATM switches with respect to wormhole routing switches in order to provide a comparable throughput to TCP connections. By contrast, ATM is more capable to handle severe congestion situations and to provide better fairness.

An efficient implementation of tree-based multicast routing for distributed shared-memory multiprocessors

This paper presents an efficient routing and flow control mechanism to implement multidestination message passing in wormhole networks. The mechanism is a variation of tree-based multicast with pruning to recover from deadlocks and it is well suited for distributed shared-memory multiprocessors (DSMs) with hardware cache coherence. It does not require any preprocessing of multicast messages reducing notably the software overhead required to send a multicast message. Also, it allows messages to use any deadlock-free routing function. The new scheme has been evaluated by simulation using synthetic loads. It achieves multicast latency reductions of 30% on average. Also it was compared with other multicast mechanisms proving its benefits. Finally, it can be easily implemented in hardware with minimal changes to existing unicast wormhole routers.

Dynamic reconfiguration of node location in wormhole networks

Abstract Several techniques have been developed to increase the performance of parallel computers. Reconfigurable networks can be used as an alternative to increase the performance. Network reconfiguration can be carried out in different ways. Our research has focused on distributed memory systems with dynamic reconfiguration of node location. Briefly, this technique consists of positioning the processors in the network depending on the existing communication pattern among them, to suit the requirements of each computation. In this article, we present a dynamic reconfiguration technique for wormhole networks. We have used both a crossbar and a multistage interconnection network to implement a reconfigurable logical two-dimensional (2-D) torus topology. The reconfiguration mechanism is based on a distributed reconfiguration algorithm. The algorithm is based on a cost function that requires only local information. We discuss reconfiguration features and adjust the different parameters of the reconfiguration algorithm. We have also studied the deadlock problem in reconfigurable wormhole networks, and give details of our solution. Finally, we have evaluated the performance of this technique under several workloads.

Adaptive and deadlock-free routing for irregular faulty patterns in mesh multicomputer

Message routing achieves the internode communication in parallel computers. A reliable routing is supposed to be deadlock-free and fault-tolerant. While many routing algorithms are able to tolerate a large number of faults enclosed by rectangular faulty blocks, there is no existing algorithm that is capable of handling irregular faulty patterns for wormhole networks. In this paper, a two-staged adaptive and deadlock-free routing algorithm called "Routing for Irregular Faulty Patterns" (RIFP) is proposed. It can tolerate irregular faulty patterns by transmitting messages from sources or to destinations within faulty blocks via multiple "intermediate nodes." A method employed by RIFP is first introduced to generate intermediate nodes using the local failure information. By its aid, two communicating nodes can always exchange their data or intermediate results if there is at least one path between them. RIFP needs two virtual channels per physical link in meshes.

Flexible and efficient routing based on progressive deadlock recovery

The development of fully adaptive, cut-through (wormhole) networks is important for achieving high performance in communication-critical parallel processor systems. Increased flexibility in routing allows network bandwidth to be used efficiently, but also creates more opportunity for cyclic resource dependencies to form which can cause deadlock. If not guarded against, deadlocks in routing make packets block in the network indefinitely and, eventually, could result in the entire network coming to a complete standstill. The paper presents a simple, flexible, and efficient routing approach for multicomputer interconnection networks which is based on progressive deadlock recovery as proposed to deadlock avoidance or regressive deadlock recovery. Performance is optimized by allowing the maximum routing freedom provided by network resources to be exploited. True fully adaptive routing is supported in which all physical and virtual channels at each node in the network are available to packets without regard for deadlocks. Deadlock cycles, upon forming, are efficiently broken in finite time by progressively routing one of the blocked packets through a connected, deadlock-free recovery path. This routing approach enables the design of high-throughput networks that provide excellent performance. Simulations indicate that progressive deadlock recovery routing can improve throughput by as much as 45 percent and 25 percent over leading deadlock avoidance-based and regressive recovery-based routing schemes, respectively.

Throttle and preempt: A flow control policy for real-time traffic in wormhole networks

Abstract In this paper, we study wormhole routed networks and envision their suitability for real-time traffic in a priority-driven paradigm. A traditional blocking flow control in wormhole routing may lead to a priority inversion in the sense that high priority packets are blocked by low priority packets for unlimited time. This uncontrolled priority inversion causes the frequent deadline missing even at a low network load. This paper therefore proposes a new flow control called throttle and preempt flow control, where high priority packets can preempt network resources held by low priority packets, if necessary. As a result, this flow control does not cause priority inversion. Our simulations show that the throttle and preempt flow control dramatically reduces deadline miss ratio for various real-time traffic configurations without extra virtual channels. It is also observed that the throttle and preempt flow control offers shorter delay for non-real-time traffic than the existing real-time flow control does.