Global Navigation Satellite Systems (GNSS) are crucial for modern infrastructure but are vulnerable to jamming due to their open structure, limited security, narrow bandwidth, and weakened signal strength over long distances. Jamming attacks exploit these vulnerabilities, disrupting both civilian and military operations by degrading or blocking positioning, navigation, and timing (PNT) services. Existing jamming detection methods rely on power-hungry Deep Learning (DL) models built with Artificial Neural Networks (ANNs) and pre-trained architectures, making them unsuitable for power- and resource-constrained edge devices used in GNSS systems and are prone to errors in noisy environments. To address these challenges, we propose leveraging Convolutional Spiking Neural Networks (CSNNs) for jamming classification, taking advantage of the energy-efficient, asynchronous, and sparse nature of SNNs. We integrate attention mechanisms and multi-axis data augmentation to enhance classification accuracy and mitigate overfitting. Our approach improves radio environment awareness in wireless communication systems by providing superior jamming classification accuracy and robust physical-layer security, particularly in noisy environments. Experimental results show that CSNN-based models achieve competitive accuracies of up to 98%, comparable to CNN-CBAM and ResNet, while maintaining accuracies as high as 97% under multipath noise with superior computational efficiency, where traditional models degrade to as low as 33%.

Delay-Tolerant Networks (DTNs) encounter unique challenges arising from intermittent connectivity, which poses reliable communication as a vital yet elusive objective. Space networks are scheduled DTNs, which facilitate the design of a routing approach that exploits knowledge of predicted contacts. However, the reliability of these contacts may deteriorate due to various factors, particularly as the network expands both in size and operational complexity. This can result in buffering levels and losses that exceed expectations. We model a DTN overlay link that is affected by unreliable contacts and limited buffer capacity by using a Continuous Time Markov Chain. Following validation against simulation results, the model is applied to the analysis of common routing strategies that consider or ignore information about the reliability of contacts. This model reveals the key performance weaknesses within these strategies. Next, we discuss two potential heuristics aimed at optimizing bundle throughput and traffic distribution. The study, therefore, provides a theoretical evaluation framework for routing strategies in DTNs that may be affected by unreliable contacts and offers insights into the achievable bundle performance under less-than-ideal conditions, which can help further advance DTN routing research.

Age of information (AoI) is a metric that measures the time elapsed since the last received update was generated and gives a measure of the timeliness of information updates in a network. In this study, we introduce an automated buffer management strategy that leverages machine learning to minimize the average AoI (or Peak AoI). This method determines the optimal stochastic policy, accounting for factors like network congestion and flow features. It relies on a Graph Convolutional Network (GCN) to determine the optimal message drop probability. The training process uses Proximal Policy Optimization with cyclic average observations of network and flow features, along with associated rewards. Specifically, the policy reward is based on the negative average Peak Age of Information (AoI) observed over multiple transmission cycles. To evaluate this approach, we conducted a simulation study where the message flow competes for network resources with secondary flows. The results highlight the significant improvements achievable in AoI metrics through the proposed smart buffer management technique.

The development of integrated satellite-terrestrial networks has gained significant attention from both industry and academia in recent years, owing to their potential for delivering low latency, high dependability, strong resilience, ubiquitous connectivity and global broadband coverage services. However, due to the ever-changing nature of satellite topology and the complexity of diverse integrated satellite-terrestrial networks, routing requests is challenging. In this paper, the vehicle movement is uncertain introducing the intermittent connectivity related to vehicles. Therefore, we propose a distributionally robust optimization (DRO) model to minimize, under uncertain latency probability distributions, the expected worst-case overall task routing delay from source to target user equipment through satellite constellation. The model addresses undetermined uploading and downloading latency between automobiles, satellites, and user equipment by employing the Wasserstein ambiguity set, allowing for unpredictable vehicle mobility and intermittent connections. By reformulating the problem into a tractable form, we determine the optimal routing path for task uploading, satellite constellation, and task downloading. Ultimately, the performance of the proposed DRO model demonstrates the model’s ability to address the challenges of integrated satellite-terrestrial network routing.

Due to the ability to provide low latency, high dependability, and worldwide broadband coverage services, the development of integrated satellite-terrestrial networks has attracted significant interest from both industry and academia over the past few decades. However, the dynamic satellite topology, heterogeneous, expansive, and intricate properties of the integrated satellite-terrestrial network make routing tasks difficult. In this research, we design the distributionally robust optimization (DRO) model with the objective of minimizing the estimated worst-case total task routing delay from the source mobile devices to the matching target mobile devices under an uncertain probability distribution. Taking into account the unpredictable vehicle movement and discontinuous connection between vehicles and mobile devices, the indeterminate offloading and downloading from automobiles to satellites and mobile devices, respectively, are captured by the Wasserstein ambiguity set. Then, we are able to determine the optimal route for task uploading, routing throughout the satellite constellation, and downloading. Finally, experimental results demonstrate that our proposed model has a lower and more robust latency than that of robust optimization (RO) strategy.

In addition to the disconnected and dynamically changing graph that characterizes a wireless delay-tolerant network (DTN), the constrained storage capacity of the nodes can lead to buffer overflows. DTN routing is commonly designed to either minimize the end-to-end data forwarding latency or maximize throughput without paying much attention to the close interrelation between the two metrics. When used in isolation, results may be conflicting, e.g., achieve low latency by routing bundles through paths with a high drop rate, that is, low throughput. A comparative analysis of the theoretical performance of a latency-only, inverse throughput-only, and combined routing objective is provided by modeling the system with a finite Continuous Time Markov Chain. The Pareto front of the multi-objective latency-throughput routing case is discussed and approached with the inverse of Kleinrock's power metric and a simplified variant. Moreover, the conditions for a latency-throughput paradox in a DTN are discussed where the regular latency trend with higher traffic loads is reversed. The study brings new insight into the multi-objective DTN routing problem and implications in the design of robust path selection methods.

Recent advances in satellite technology and the introduction of the Space-Data-as-a-Service business models are attracting interest for using the platform in a wide range of applications that include earth observations, 5G integration, aeronautical and maritime tracking and communications, asset tracking, sensor data collection, among others. In this new scenario, it becomes of interest not only selecting a ground station provider based on the monetary cost, but also the download features with prevailing atmospheric conditions to achieve high reliability and low data delivery latency. To this end, the Cognitive Space Gateway approach is used to route data bundles in a Ground Station-as-a-Service context with multiple providers to find a suitable balance between data latency and the ground station monetary costs, exploiting as possible site-diversity when available. The approach is tested experimentally using NASA's High-Rate Delay-Tolerant Networking (HDTN).

Delay-tolerant networking (DTN) provides the mechanisms needed for both single and multihop space communications when cyclical bandwidth is available between the nodes. Because node contacts can be infrequent, achieving high bandwidth utilization during contacts or extending their duration as possible are topics of special relevance. In this work, an online approach for the Licklider Transmission Protocol (LTP) that is suitable for small size, weight, and power (SWaP) devices is suggested. The method dynamically searches for the optimal segmentation, i.e., the maximum payload size to be used to transmit data blocks. It is contrasted with a Q-learning-based method, which achieves similar performance with larger space complexity. Either option attains reliable bundle transmissions with acceptable delay and throughput levels well beyond the range achievable with static payload lengths and adverse conditions of high loss rates. The approach has minimal information requirements and does not require knowledge of the current bit-error-rate, transmission rate, modulation type, power, gains, nor the one-way light times. The method also requires minimal changes to the sending engine and is fully compatible with standard LTP receive engines. The technique can help to improve the effective DTN capacity by extending the usable contact times and achieving transmissions with improved performance when signal fluctuations occur.

This paper discusses changes to the Cognitive Space Gateway (CSG) approach to bundle routing to make it suitable for use with standard delay-tolerant networking (DTN) protocols and the High Rate DTN (HDTN) architecture. The revision removes the current convergence-layer information dependency of the CSG with an alternative mechanism that entails the exchange of performance-related data between adjacent nodes along the bundle path using extended canonical blocks. The proposed approach allows estimating learning rewards for the reinforcement learning method of the CSG using standard HDTN services. Possible changes to the HDTN architecture are also discussed to enable the implementation of dynamic routing decisions for bundles. The revised mechanisms are experimentally tested and the results both verify the suitability of the changes and offer a practical demonstration of cognitive routing capabilities with HDTN.

With the explosive growth of low earth orbit (LEO) satellite networks, such as Starlink, satellite communication has lower latency and can achieve high-speed transmission than before. However, the time-variant topology during all network lifetimes makes the routing problem in the LEO satellite networks challenging. Therefore, in this paper, we propose the deep reinforcement learning-based satellite routing (DRL-SR) method to tackle the multi-commodity flow routing problem in the LEO satellite networks. Given the current state of the satellite network environment, the satellite operation center will determine how to route the requests to the matching destinations. Particularly, the single agent in our DRL-SR approach can determine the multiple next hops as actions for all the corresponding requests each timeslot. Finally, simulation results show that our proposed algorithm yields lower latency than the shortest path approach.

For a class of delay-tolerant networks, it is assumed that the future topology changes are predictable and therefore, it is possible to find paths for data bundles by traversing the contact graph. This is the method used by Schedule-Aware Bundle Routing through Contact Graph Routing and the reward shaping method proposed for the Cognitive Space Gateway (CSG). In this work, the performance impact of anticipated contacts that randomly fail to occur is investigated. A queuing model is first formulated to gain insight into the end-to-end bundle delivery performance drop compared to the optimal performance. Then, experimental results provide further indication of the performance impact using an emulated space DTN environment. While a contact miss event can be hard to predict or to avoid, bundle re-routing is investigated for CSG routing as a possible mitigation method and is experimentally tested.

The results from an experimental evaluation study of cognitive network routing applied to the transmission of a scientific dataset and assuming a delay-tolerant space communications context are reported in this paper. The routing method that was tested relies on a spiking neural network (SNN) both to encode network performance predictions and to select the optimal outbound link for each data bundle. By defining an adaptive approach to bundle routing, the method can cope with changes in both the network congestion levels and the availability of redundant paths. Repeatable experiments were carried out using a laboratory network testbed to characterize the average routing performance. The reference implementation of Contact Graph Routing from the Interplanetary Overlay Network (ION) distribution was used to obtain the baseline performance of the current state of practice in space network routing. The results demonstrate the performance advantages of the cognitive networking approach to bundle routing when the network is being affected by different kinds of impairments, such as channel losses and link disruptions.

Delay tolerant networks are challenged networks that are characterized by intermittent connectivity that randomly impacts end-to-end performance. One way to improve the quality of communications is by routing data (i.e., bundles) through more than one channel or end-to-end path whenever the case permits. We analyze the probabilistic optimal routing problem for this case through the formulation of a queuing model with vacations that serve to represent the transmission of data bundles over channels of sporadic availability. Despite transmissions could be scheduled in some cases (e.g., knowledge of the expected contact plan can be known in space networks) we assume unpredictable transmission opportunities where it is not possible to establish the sequence of future contacts beforehand. This makes the analysis applicable to different kinds of delay tolerant networks. For example, with the Licklider Transmission Protocol, when a disruption occurs in the midst of a bundle transmission, the service can be temporarily stopped and restarted at the next contact opportunity. A numerical solution of the model illustrates the trade-offs that exist in the optimization of bundle flows over multiple space channels.

The Cognitive Network Controller (CNC) defines a neuromorphic architecture where a spiking neural network can both encode network performance observations and select the optimal actions (e.g., routes) for the context of those observations. Because of these features, the CNC can quickly adapt to changes in the operational environment to either maintain or improve selected performance metrics. This behavior can be attractive for a space networking scenario with orbiting and ground-based assets that are either stationary or manned, bringing an elevated level of autonomy in network communication decisions. Using the SCaN testbed as a laboratory facility in orbit, we evaluated the adaptation abilities of the CNC applied to a space network routing application. Towards this end, the CNC design and the related neuromorphic processor were implemented in software and deployed on the flight computer of the SCaN testbed, and then applied to route bundles to a ground station over parallel links. This work likely constitutes the earliest demonstration of a space application for neuromorphic computing and a basic validation of the online adaptation capabilities of the CNC.

In this paper, results from an experimental study of the end-to-end bundle delivery performance of two concurrent flows transmitted over a delay-tolerant network are presented. The study compares the Contact Graph Routing algorithm, which is the foundation of the CCSDS standard Schedule Aware Bundle Routing and the Cognitive Space Gateway that has been recently introduced as a cognitive networking alternative to the problem of space bundle routing. Both algorithms make non-cooperative, dynamic routing decisions for bundles at each step based on a similar utility, which is formulated as the minimum expected bundle delivery time to the destination. The study aims to find how well each algorithm plays the routing game under different conditions. The experiments were carried out on a laboratory testbed with emulated Earth-Moon communication conditions and included evaluation cases with a permanently connected substrate and a substrate that is being affected by regular link disruptions. Also, the flows were evaluated with cases where their shortest path either overlap or not. The results indicate that both routing approaches play a coherent game achieving fair performance for both flows, but with the CSG achieving better performance than the CGR approach.

The time required to reliably deliver a data block with the Licklider Transmission Protocol (LTP) depends on the frame losses occurring during the transmission. LTP establishes overlay links that involve one or more transmission sections at the underlay. Because the state of these sections is time-dependent, LTP works without specific knowledge of the underlay. This property creates challenges to the block segmentation as it involves a tradeoff between the overall header overhead and the higher loss rates of large segments. In this paper, a practical segment loss mitigation method is proposed that benefits block delivery times by deciding segment lengths prior to each block transmission. This goal is achieved with a cognitive networking approach to the problem that leverages the parallel processing and storage capabilities of neuromorphic computing, which is prospectively adequate for onboard systems of known constraints in size and electrical power. The key advantage of this online method is that it does not need complete information about the channel properties, models, protocols, nor state. Instead, the method learns autonomously the best segment length for the current conditions by mapping the delivery performance of prior blocks to the synapse strengths of a spiking neural network, which is then used to generate new decisions for the next segments. Simulation results provide an indication of the performance benefits.

We introduce a neuromorphic architecture that is suitable for computing the next-hop for bundles in a disruption or delay-tolerant network. We define gateways that use spiking neural networks (SNN) that learn how to make autonomous routing decisions by observing single-hop bundle transmission performance and without requiring knowledge of the global network state. To formulate learning rewards, the gateways predict the end-to-end response time using local information. The rewards are then absorbed into the synapse strengths of the SNNs which estimate the relative Q-values of the next-hop alternatives for the next bundles. The use of the SNNs leads to improved routing performance with respect to performance goals of interest. Unlike comparable routing techniques, the proposed method achieves traffic balancing when multiple outbound links are available to send bundles to their next hop. An experimental evaluation of a proof-of-concept confirms that the proposed architecture is able to achieve high performance despite the presence of different kinds of communication impairments, such as frequent and lengthy link disruptions, long propagation delays and erratic transmission performance.

A delay-tolerant network (DTN) is characterized by an intermittent topology where nodes and links may be down for long periods of time. This work evaluates whether it is more efficient to stop and resume bundle transmissions on the next contact opportunity after they become interrupted by an ending contact (preemptive model) than dropping the affected bundles (lossy model). Common intuition tells that the former model involves higher response times than the latter, but also fewer losses, so the optimal choice may not be immediately evident. For the analysis, stochastic models that describe the bundle transmission process over a DTN channel that is characterized by recurrent and random contact opportunities are formulated. Simulation results confirm the accuracy of the models. The study reveals the performance trade-offs between the two design choices and indicates that, contrary to expectations, the lossy model makes more efficient use of the channel than with preemption according to Kleinrock's power metric, which integrates throughput, delay, and loss. However, after considering retransmissions, preemption yields slightly higher power than the lossy approach. This work brings new insights into DTN performance and design choices.

We consider a dynamic reservoir of spiking neurons to generate routing decisions for bundle transmissions in a delay tolerant network (DTN). The reservoir transforms the context, as described by selected system metrics, and the temporal information that is contained in the network's contact list into a high-dimensional feature space, which allows quickly determining the next-hop for bundles whenever needed. The method can be of particular interest for space networks, which are characterized by resource-constrained nodes, known contact opportunities, and dynamic features, such as the state of the channels and buffers. The learning process occurs offline using linear regression and involves a modest computational effort. After training, the spiking neural network can be used to achieve an efficient network operation, allowing to substitute the costly per-bundle shortest path computation with a simple "recall" action of the best next-hop for the current conditions as stored in the reservoir. We investigate different methods to translate the contact list information into a suitable input for the reservoir and evaluate the approach considering the impact of the readout function size applied to GEO-LEO DTN scenarios.

Real-time electricity prices and volatile workload patterns have a major impact on the operational costs of distributed data centers. A smart grid enables large consumers to regulate their usage patterns with the knowledge of real-time prices at different locations, and their energy consumption through the demand response (DR) program. However, the price inclined workload assignment may lead to an increase in peak demand, and decrease the benefits of workload offloading. Moreover, energy-efficient service management is challenged by unpredictable network conditions that can cause service performance degradation. Using an actor-critic approach, we developed a system that distributes a service's load among distributed sites to take advantage of the spatial variations in energy pricing. Our experimental results on the CloudLab testbed have proved with 45% reduction in running costs without impacting the response time using the Wikipedia request traces considering peak-based pricing.

We discuss the online optimization of redundant copies of computing tasks. The method helps to overcome the limitations of cloud and edge computing methods for mobile applications where the shared use of mixed and distributed processors often yields unpredictable execution times. While the aim is to obtain the earliest response (discarding the rest), without careful control of the task replication process, the redundant executions may lead to excessive processor contention leading to undesired effects. The current state of practice assumes the use of homogeneous processors, heuristics, and perfect knowledge of the future task runtimes, which limit the application scope of the idea. Through reinforcement learning, the proposed method discovers autonomously how to optimally select for each new request both the number of replicas and their processor assignment. The method can operate effectively without full knowledge of system features and regardless of the changing system state. An extensive simulation study using different scenarios confirms the performance of this proposal and offers a quantitative insight into its advantages and limitations. Diverse mobile applications prospectively benefit from this approach, as computer and communication networks become larger and more diverse, and mobile applications require increasingly higher reliability and lower latency.

The Licklider transmission protocol is a point-to-point communication protocol designed for space links, which commonly involve extreme delays, disruptions, and lossy transmissions. The protocol sends application data in blocks, which in turn are sent in segments. It achieves reliable block delivery through multiple transmission rounds, each one re-sending the segments lost during the previous round. This retransmission process drives protocol performance. We derive exact and approximate methods to find the average number of rounds per block. Then, we estimate the block delivery time and other metrics using this value. We found that the common practice of matching segment lengths to the maximum transfer unit of the link layer may lead to suboptimal performance. The models provide accurate protocol performance prediction, which can help to optimize protocol parameters for specified operating conditions.

Cognitive networks plan, decide, and act at different layers of the protocol stack, based on perceived conditions of the network state and assigned rules. We introduce a Cognitive Network Controller (CNC) that relies on a spiking neural network to implement artificial cognition. Based on the outcome of prior actions, the CNC learns how to improve future decisions to maximize its assigned objective. We present the design of the CNC and associated spiking neural network model, a decision encoding strategy based on the time-to- fire of spikes, a learning rule, and synapse weight management. Then, we evaluate the effectiveness of the proposed method using an event-driven simulation of the online selection of end-to-end paths for file transfers over a network of mixed performance. The results indicate that the CNC effectively learns how to dynamically select paths for file transfers that produce lower latency than conventional methods for a broad range of network conditions. The proposed CNC potentially facilitates the integration of cognition and future network applications that require autonomous performance optimization.

Despite the notable progress that has been achieved on long-haul radio and optical communications, which are expected to improve the data transfer rates of future space exploration missions, both scheduled and random link outages remain possible. To improve reliability, NASA's delay-tolerant networking (DTN) architecture offers the Bundle Protocol (BP) and the Licklider Transmission Protocol (LTP) that implement bundle custodial transfer and automatic repeat query retransmissions. We analyze the performance of these protocols considering the default flow control method of LTP, which limits the number of concurrent sessions. We derive the combined BP and LTP throughput and provide approximations for model parameters that are difficult to find exactly. Event-driven simulations suggest that our analysis is reasonably accurate under different channel conditions and protocol settings. The proposed model provides insight into the role of the channel conditions, the workload, and the choice of BPLTP protocol parameters on the performance of a reliable space link.

The web infrastructure continues to grow in both size and diversity and is quickly approaching 2 billion sites that serve an equally massive number of users. Content replication is a common technique that aims to improve reliability and performance. It consists of the use of multiple server mirrors that operate across geographically distributed regions. Several factors impair the static system optimization or even a periodic optimization. Final energy costs depend on the level of workload handled by each region given local variations in energy pricing. To accentuate the problem, energy pricing may dynamically change for some sites. Also, network state fluctuations, that are produced by congestion and failures, create time-varying performance to different users. These factors contribute to affect user experience and service costs. We propose an energy-delay aware web routing method that dynamically directs user requests to a set of replicated sites. The method relies on learning to implement a customized performance-cost load-balancing. We present experimental results from a network testbed using actual power measurements and simulated spatial variations in the energy prices according to the U.S. electricity market. The results show that the method can help to achieve the desired balance of energy cost and average response time, while reducing energy costs up to 11% without introducing a major impact to service quality.

Reverse proxies are key performance elements of a content distribution network (CDN) and operate as intermediaries between client and upstream (or origin) server traffic. A typical reverse proxy server aims to distribute user workload, more or less equally, among multiple upstream servers. To keep the pace with the growing demand for content, CDNs regularly expand their capacity by adding extra upstream servers to handle the increased traffic. However, distributing traffic without consideration of the energy and performance features of the machines and keeping servers active all the time can reduce energy efficiency drastically because of the seasonal variation of traffic. This outcome is highly undesirable because of environmental and economic concerns. Policy-based management is an approach for large-scale network management that helps to simplify network administration through the use of event-action rules. We investigate the formulation of server selection policies for reverse proxies to improve the energy efficiency of the system while maintaining desired service-level objectives. In particular, we formulate four energy-efficient policies of varying complexity and performance, and evaluate them using realistic data obtained from a testbed. The results are also contrasted with the outcome of common load distribution methods that ship with a popular open-source proxy server to show the advantages of the proposed policy generation strategies.

Virtualization allows multiple applications to run on different execution platforms, but sharing the same host machine. A better knowledge of the expected power consumption of computer hosts that run virtualized applications could help to improve capacity planning and optimization of cloud systems that use virtualization for resource management. In this paper, power and performance predictions are estimated from utilization figures of the main computer subsystems (CPU cores, drives, memory, and network ports), which handle the aggregated tasks produced by the virtualized applications. Extensive measurements conducted on two different systems validate the model.

The widespread of mobile devices, such as Wi-Fi enabled laptops, PDAs and smart phones, makes mobile ad hoc networks (MANET) an attractive and low-cost alternative to create extended network service areas for users on the move. Unfortunately, the establishment of reliable multi-hop routes with mobile nodes that forward packets is a difficult task because of the temporal availability of links. In this paper, we revisit and evaluate the link durability routing protocol (LDRP), a low-overhead MANET protocol that can make use of node localization to improve the reliability of routes. We evaluate the protocol on a hybrid MANET/sensor network system where sensors provide tracking services to mobiles and the information needed by LDRP to achieve improved routing. In particular, we investigate the level of MANET reliability that can be achieved with sensor networks offering different levels of localization accuracy by altering network parameters, such as sensor density and communication patterns. A simulation study quantifies the benefits of the approach over AODV.

We describe a prototype of trading system platform populated by agents who autonomously decide to buy and/or sell items according to a set of local needs which arise dynamically (also by possibly accessing information provided by pervasive devices) by in the process of fulfilling a given overall utility. The market has combinatorial nature in a way that items to be traded are combined into packages, in accordance with a principle that drives the nature of many current markets. However, differently from these, items belong to a number of distinct sellers distributed in the platform, and are chosen singularly on the basis of buyers preferences and needs. Agents are thus situation-aware, with sellers coming acquainted of the market demand, and buyers price offers, through a Knowledge Network. This latter drives the way market offers balance the demand by gathering the needed information in an autonomous way and taking advantage of pervasive devices. Packaging is realized by agent aggregation into Virtual Sellers, in an autonomous fashion, and we propose an opportunistic policy whereby aggregation is governed by a Combinatorial Auction. The market is studied through proof-of-concept simulation, where the efficiency deriving from the opportunistic aggregation based on Combinatorial Auctions and the influence of contextual self-awareness are studied.

The main activities leading towards a testbed validation of routing protocols for MANETs involve code development and experimentation. Because of the characteristics of such networks, these activities are commonly obstructed by the need of managing parameters inherently difficult to control in real situations, such as the position of the nodes or the starting residual energy of mobiles. This paper discusses a possible solution to the problem and presents an implementation of the proposal. The MANET testbed manager (MTM) is a system that allows setting up reproducible testbed scenarios useful for code development, initial validation and visualization. The main purpose of MTM is to simulate the main physical elements that can be found in a mobile network while allowing a MANET routing protocol operate concurrently being unaware of the underlying simulation.

This paper addresses two scalability problems related to the cognitive map of packets in ad hoc cognitive packet networks and proposes a solution. Previous works have included latency as part of the routing goal of smart packets, which requires packets to collect their arrival time at each node in a path. Such a requirement resulted in a packet overhead proportional to the path length. The second problem is that the multiplicative form of path availability, which was employed to measure resources, loses accuracy in long paths. To solve these problems, new goals are proposed in this paper. These goals are linear functions of low-overhead metrics and can provide similar performance results with lower cost. One direct result shown in simulation is that smart packets driven by a linear function of path length and buffer occupancy can effectively balance the traffic of multiple flows without the large overhead that would be needed if round-trip delay was used. In addition, energy-aware routing is also studied under this scheme as well as link selection based on their expected level of security.

An standard testing framework is needed to satisfactorily validate mobile ad hoc network (MANET) protocol implementations and carry out comparison studies of different approaches. Research groups have employed dissimilar methodologies in experimentation and no common testing framework is available to the MANET research community. Furthermore, testbed experimentation faces the challenge of creating reproducible scenarios that are not only expensive in use of resources but also difficult to achieve. This paper presents MANET testbed manager (MTM), a system for setting up testbed scenarios for MANET protocols. MTM makes available an integrated framework for control, experimentation and visualization and relies on emulation of elements of a wireless mobile network that are difficult to manipulate, such as the position of the nodes, mobility patterns and energy consumption. MTM's goal is to offer a highly-flexible and low-cost testing environment for protocol validation to fill the gap between simulation and testbed experimentation.

Novel user-oriented networked systems will simultaneously exploit a variety of wired and wireless communication modalities to offer different levels of quality of service (QoS), including reliability and security to users, low economic cost, and performance. Within a single such user-oriented network, different connections themselves may differ from each other with respect to QoS needs. Similarly, the communication infrastructure used by such a network will, in general, be shared among many different networks and users so that the resources available will fluctuate over time, both on the long and short term. Such a user-oriented network will not usually have precise information about the infrastructure it is using at any given instant of time, so that its knowledge should be acquired from online observations. Thus, we suggest that user-oriented networks should exploit self-adaptiveness to try to obtain the best possible QoS for all their connections. In this paper we review experiments which illustrate how "self-awareness," through online self-monitoring and measurement, coupled with intelligent adaptive behavior in response to observations, can be used to offer user-oriented QoS. Our presentation is based on ongoing experimental work with several "cognitive packet network" testbeds that we have developed.

Based on our earlier work ("Towards networks with intelligent packets", Proc. 14th Int. Symp. on Computer and Information Sciences, p. 1-11, Oct. 1999), we discuss packet networks in which intelligent capabilities for routing and flow control are concentrated in the packets, rather than in the nodes and protocols. This paper describes a possible testbed to test and evaluate their capabilities, and presents an analytical model for the worst and best case performance of such systems.