Fig 7 - uploaded by Roberto Rigolin F. Lopes
Content may be subject to copyright.
Source publication
This paper describes a hierarchical queuing mechanism designed to monitor and control radio buffers while delivering web services in tactical networks. Our solution was motivated by the restrictions imposed by VHF radios, which have large coverage (∼20 km) but very low data rate (e.g. up to 9.6 kbps). Therefore we implemented two queues, one for me...
Contexts in source publication
Context 1
... flow diagram in Fig. 7 represents the control loop together with the core services implementing it. First, the messages are intercepted by a proxy which sends it to the message queue service together with its QoS requirements. Given the destination address, the routing service returns the next hop; then the message is queued in the Qm r for that radio r. In ...
Context 2
... Given the destination address, returns the next IP hop which is mapped to the radio's Qm; (2) UDP transport: compresses and fragments the message into UDP packets; -Packet handler: queues IP packets while monitoring the radio buffer. We implemented two ways of controlling IP packets (i and ii in Fig. 7). i) If ΔB i < b%, it sends packets (dequeue). Else, it queues the IP packets and waits until the buffer is below the threshold b%. ii) given the current occupancy, compute the number of packets to keep the buffer usage within the b% threshold; (3) Radio plug-in: pulls the radio buffer occupancy and shares with the contextual ...
Context 3
... notice that there are delays ( 1,2,3,4 in Fig. 7) while compiling/retrieving the current context (e.g. the VHF radios update their buffer occupancy every second). If for any reason the sum of these delays is higher than the time to fill up the buffer (Eq. 4), a buffer overflow will ...
Context 4
... illustrated in Fig. 1) and message size (500 kB) that overflows the VHF radios as earlier reported in subsection III-C (The problem). The VHF radios were configured to a data rate of 4.8 kbps so each execution, sending 500 kBytes, took about ∼1000 seconds (∼16 minutes). Let us plot the behavior during the three-step algorithm earlier described (Fig. 7), starting with the radio buffer usage ΔB (1), followed by the packet queue Qp size (2) and finishing with the values returned by the "How long to admission?" function (3), as ...
Context 5
... configured to deliver the radio's buffer occupancy every second because the radio plugin pulls radio status every second (it can't be faster for this particular radio). In this section we vary the b% threshold to discover both lower/upper limits of our solution; this exercise also motivated the definition of a proactive control mechanism, (ii) in Fig. ...
Citations
... In [19], we defined a hierarchy of queues to adapt the user data flows to the current network conditions to avoid radio buffer overflow. The control mechanism was reactive, pausing the data flow when a pre-defined threshold was crossed. ...
... We call it inter-packet interval (Q delay 0 ), which is computed using the cross-layer information shared by the radio (layer 0) and the routing protocol (layer 1). In previous investigations, we defined two mechanisms: (i) using the radio buffer occupancy [19] and (ii) using the estimated link data rate compiled by the routing protocol [2], [4]. Here, we combine these two solutions using the following system of relations. ...
... Our experiments reused the 500 kB data flow from previous investigations for comparison purposes [4], [19] and monitored both sender and receiver (highlighted earlier in Fig. 1) at the three layers, namely radio buffer (Q 0 at layer 0), queue of packets (Q 1 at layer 1) and queue of messages (Q 2 at layer 2). Table 4 summarizes the experimental setup listing the radio type, time window, buffer threshold, data flows, network conditions and routing protocol used during the experiments discussed in the next sub-sections. ...
This paper introduces a hierarchy of queues complementing each other to handle ever-changing communication scenarios in tactical networks. The first queue stores the QoS-constrained messages from command and control systems. These messages are fragmented into IP packets, which are stored in a queue of packets (second) to be sent to the radio buffer (third), which is a queue with limited space therefore, open to overflow. We start with the hypothesis that these three queues can handle ever-changing user(s) data flows (problem A) through ever-changing network conditions (problem B) using cross-layer information exchange, such as buffer occupancy, data rate, queue size and latency (problem A|B). We introduce two stochastic models to create sequences of QoS-constrained messages (A) and to create ever-changing network conditions (B). In sequence, we sketch a control loop to shape A to B to test our hypothesis using model A|B, which defines enforcement points at the incoming/outgoing chains of the system together with a control plane. Then, we discuss experimental results in a network with VHF radios using data flows that overflows the radio buffer over ever-changing data rate patterns. We discuss quantitative results showing the performance and limitations of our solutions for problems A, B and A|B.
... OLSRv2 was in charge of compiling and sharing the network topology among all mobile nodes in the testbed, providing topology management with proactive neighbor discovery and calculating routes using the DAT metric. Within a Tactical Service-oriented Infrastructure (TSI [12], [23]), olsrd also shared the routes and the network topology (graph) with other services in a cross-layer orchestration using the jsoninfo plugin 3 . ...
This paper describes a quantitative evaluation of OLSRv2 proactive neighbor discovery in a heterogeneous tactical network using real military radios. Tactical networks generally have low data rates, high latency and hard to predict mobility patterns. The experimental network used in this investigation hosted mobile nodes using three different types of communication technologies, namely VHF (9.6 kbps, 20 km), UHF (240 kbps, 2 km) and SatCom (100-500 kbps-simulated). OLSRv2 was set to use multicast IPs to send hello and topology control messages within intervals suitable for the data rate available; using less than ∼5% of the network nominal capacity. Therefore, we analyze results from a testbed with mobile nodes connected to a tactical network using OLSRv2 to create an overlay IP network. The quantitative results suggest that OLSRv2 can be fine-tuned to proactively discover IP nodes in radio networks with quite different link data rates. The trade-off between data overhead and time to converge is discussed over estimated metrics.
... The goal is to assess how the recent developments in commercial technologies, particularly SDN, NFV and NSF, can facilitate the deployment of multi-nation military networks. The military has several specific requirements that are different from the civilian communication systems [7]- [9]. ...
Modern multi-nation military communication and information systems demand rapid deployment and reconfiguration to enable secure information exchange between domains belonging to different nations. In order to support such a multi-nation mission scenarios NATO has been developing the concept of a federated mission network (FMN). These networks are leveraging the advantages of software-defined networking (SDN) and network function virtualization (NFV) to adapt to a wide range of security requirements using network security functions (NSF). The investigation reported in this paper discusses two use cases, namely how to automate security policy evaluation, and how to deploy a security guard (Information Exchange Function) between network segments with different classification. Both cases use network scenarios from protected core network (PCN) because the goal is to compile open challenges to automate the deployment and management of secure coalition networks.
... Military systems have to manage the limitations of the radio devices, such as link data rate and buffer size. We noticed that the limited buffer size combined with the ever-changing network scenario and user data flow (say message size and time distribution) results in a buffer overflow [29]. This fact imposes a restriction on the range of experiments TNT could support. ...
This paper addresses the challenge of testing military systems and applications over different communication scenarios with both network conditions and user data flows changing independently. We assume that systems developed to handle ever-changing communication scenarios are more likely to be reliable and robust during real military operations. Therefore, we propose the Tactical Network Test (TNT) platform to automate the evaluation of military systems and applications over real military radios using a reproducible test methodology. TNT has four main goals (i) the creation of QoS-constrained data flows; (ii) the execution of models to change network conditions; (iii) a performance evaluation of a military system over ever-changing conditions; and (iv) the monitoring and performing data analysis. Our platform was used to execute experiments in a VHF network by sending uniformly distributed data flows during seven different communication scenarios, either generated by a stochastic model or mobility models. The experimental results are used to discuss the military system's performance by quantitative analysis using network metrics and the test scenario characterization through mobility metrics.
... Military systems have to manage the limitations of the radio devices, such as link data rate and buffer size. We noticed that the limited buffer size combined with the ever-changing network scenario and user data flow (say message size and time distribution) results in a buffer overflow [29]. This fact imposes a restriction on the range of experiments TNT could support. ...
This paper addresses the challenge of testing military systems and applications over different communication scenarios with both network conditions and user data flows changing independently. We assume that systems developed to handle ever-changing communication scenarios are more likely to be reliable and robust during real military operations. Therefore, we propose the Tactical Network Test (TNT) platform to automate the evaluation of military systems and applications over real military radios using a reproducible test methodology. TNT has four main goals (i) the creation of QoS-constrained data flows; (ii) the execution of models to change network conditions; (iii) a performance evaluation of a military system over ever-changing conditions; and (iv) the monitoring and performing data analysis. Our platform was used to execute experiments in a VHF network by sending uniformly distributed data flows during seven different communication scenarios, either generated by a stochastic model or mobility models. The experimental results are used to discuss the military system's performance by quantitative analysis using network metrics and the test scenario characterization through mobility metrics.
... We have been developing a tactical middleware, which implements a hierarchy of three queues complementing each other relying on cross-layer information exchange, as illustrated in Fig. 1 [10]. In this figure, the hierarchical data pipeline starts from message queue (1), followed by the packet queue (2) and finally to the radio buffer (3); the cross-layer information exchange is supported by a contextual monitoring service (4) which serves as an interface to the radio network. ...
... Contextual monitoring gathers the network topology and buffer occupancy details. Lastly, the radio plug-in shares the buffer occupancy with the contextual monitoring, which Ever-changing data rate patterns Fig. 1: Hierarchy of queues and core services [10] communicates to the packet handler. Notice that, implementation of Transmission Control Protocol (TCP) adds on to large recovery time in case of bandwidth-constrained and high latency networks such as VHF or HF networks. ...
... We supported with experimental results that it adapts to the ever-changing network conditions resulting in the better shaping of the data flow for a single threshold value of 10 (%). However, in those previous works [10], [11] an exploratory validation was missed, opening the opportunity to conduct an exploratory study here to a) compare the Threshold Shaping with IPI 1 for different values of threshold and b) discuss potential improvement in the computation of IPI 1 using cross-layer metrics. ...
The adaptation of tactical systems to either message outbursts of user-behavior or unpredictable network changes is a necessary condition to guarantee the robustness of the system. However, both events are likely to occur simultaneously in military scenarios, in case of which the system can use multi-layer control loops to adapt its store-and-forward mechanism. This paper is a part of a wide investigation to evaluate the robustness and solve the related issues of tactical systems. Our architecture implements the store-and-forward mechanism with a hierarchical queuing model of the message, IP packet, and radio buffer. In this study, we want to solve the data overflow in memory-constrained devices such as radio buffer caused due to data outbursts, eventually, leading to packet loss. In order to mitigate radio buffer overflow, we introduce a control point to de-queue the packets from the packet handler to the radio buffer. Our system adds an Inter-Packet-Interval (IPI) to control the admission into the radio buffer. IPI is computed using the nominal link data rate compiled by the radio together with buffer metrics such as the occupancy and threshold. We discuss experimental results from a VHF network suggesting that our cross-layer IPI allows better control of the buffer usage thereby increasing the system robustness to ever-changing link data rates in the network.
... buffer occupancy, link data rate, latency, packet loss and so on). In our previous investigations, [2], [4] and [19], we discussed simulated and experimental results testing different ways of computing the Inter-Packet Interval (IPI) using cross-layer information to adapt the queuing mechanism to the current network conditions. Here, we study the performance of a hybrid solution (reactive/proactive) using Very High Frequency (VHF) radios by challenging our mechanism with ever-changing communication scenarios including network disconnections. ...
... In [19], we defined a hierarchy of queues to adapt the user data flows to the current network conditions to avoid radio buffer overflow. The control mechanism was reactive, pausing the data flow when a pre-defined threshold was crossed. ...
... We call it inter-packet interval (Q delay 0 ), which is computed using the cross-layer information shared by the radio (layer 0) and the routing protocol (layer 1). In previous investigations, we defined two mechanisms: (i) using the radio buffer occupancy [19] and (ii) using the estimated link data rate compiled by the routing protocol [2], [4]. Here, we combine these two solutions using the following system of relations. ...
This paper introduces a hierarchy of queues complementing each other to handle ever-changing communication scenarios in tactical networks. The first queue stores the QoS-constrained messages from command and control systems. These messages are fragmented into IP packets, which are stored in a queue of packets (second) to be sent to the radio buffer (third), which is a queue with limited space therefore, open to overflow. We start with the hypothesis that these three queues can handle ever-changing user(s) data flows (problem A) through ever-changing network conditions (problem B) using cross-layer information exchange, such as buffer occupancy, data rate, queue size and latency (problem A|B). We introduce two stochastic models to create sequences of QoS-constrained messages (A) and to create ever-changing network conditions (B). In sequence, we sketch a control loop to shape A to B to test our hypothesis using model A|B, which defines enforcement points at the incoming/outgoing chains of the system together with a control plane. Then, we discuss experimental results in a network with VHF radios using data flows that overflows the radio buffer over ever-changing data rate patterns. We discuss quantitative results showing the performance and limitations of our solutions for problems A, B and A|B.
... buffer occupancy, link data rate, latency, packet loss and so on). In our previous investigations, [2], [4] and [19], we discussed simulated and experimental results testing different ways of computing the Inter-Packet Interval (IPI) using cross-layer information to adapt the queuing mechanism to the current network conditions. Here, we study the performance of a hybrid solution (reactive/proactive) using Very High Frequency (VHF) radios by challenging our mechanism with ever-changing communication scenarios including network disconnections. ...
... In [19], we defined a hierarchy of queues to adapt the user data flows to the current network conditions to avoid radio buffer overflow. The control mechanism was reactive, pausing the data flow when a pre-defined threshold was crossed. ...
... We call it inter-packet interval (Q delay 0 ), which is computed using the cross-layer information shared by the radio (layer 0) and the routing protocol (layer 1). In previous investigations, we defined two mechanisms: (i) using the radio buffer occupancy [19] and (ii) using the estimated link data rate compiled by the routing protocol [2], [4]. Here, we combine these two solutions using the following system of relations. ...
This paper introduces a hierarchy of queues complementing each other to handle ever-changing communication scenarios in tactical networks. The first queue stores the QoS-constrained messages from command and control systems. These messages are fragmented into IP packets, which are stored in a queue of packets (second) to be sent to the radio buffer (third), which is a queue with limited space therefore, open to overflow. We start with the hypothesis that these three queues can handle ever-changing user(s) data flows (problem A) through ever-changing network conditions (problem B) using cross-layer information exchange, such as buffer occupancy, data rate, queue size and latency (problem A|B). We introduce two stochastic models to create sequences of QoS-constrained messages (A) and to create ever-changing network conditions (B). In sequence, we sketch a control loop to shape A to B to test our hypothesis using model A|B, which defines enforcement points at the incoming/outgoing chains of the system together with a control plane. Then, we discuss experimental results in a network with VHF radios using data flows that overflows the radio buffer over ever-changing data rate patterns. We discuss quantitative results showing the performance and limitations of our solutions for problems A, B and A|B.
... We have been evolving a system to solve Problem A|B shaping the QoS-constrained user data-flows to the radio links. In our previous investigations, [1] and [7], we discussed simulated results testing different ways of computing the interpacket interval (IPI) using cross-layer information to adapt the queuing mechanism to the current conditions. Here, we study the performance of a hybrid solution (reactive/proactive) in VHF radios by challenging our mechanism with ever-changing data rates within a time-window. ...
... This investigation complements a previous paper [7] which discusses buffer overflow experiments in VHF radios (also PR4Gs) using a message of 500 KB, which is four times the size of the radio buffer. Here, we reuse the "big" message for comparison purposes and study the performance of our queuing mechanism over the three patterns of ever-changing data rates D 1 , D 2 and D 3 . ...
... Finally, section V concludes the paper. In our previous investigation [7], we studied the buffer overflow in VHF radios doing experiments with a message (500 KB) four times bigger than the radio buffer in a laboratory environment (radios with wired antennas) using different data rates (from 0.6 to 9.6 kbps). To address this issue, we proposed a hierarchy of queues (also in Fig. 1) to avoid the buffer overflow. ...
In this paper, we discuss experimental results testing a hierarchy of queues controlling the user data-flow over a VHF network with ever-changing data rates (up to 9.6 kbps). We challenged our solution creating three patterns of ever-changing data rates using a stochastic model to include the element of chance (randomness) that can be reproduced for quantitative comparisons. We discuss numbers showing that our queuing mechanism adapts its behavior (i.e. shaping the user data-flow) to the network conditions using feedback from the radio buffer (reactive) and from the routing protocol (proactive). Thus, our hybrid solution monitors the radio buffer occupancy to pause the transmission when a threshold is crossed, and proactively adds an inter-packet interval (IPI). The IPI varies as a function of the link data rate (computed by a tactical router), current network usage, packet loss and latency. The experimental results show three queues (for messages, IP packets and the radio buffer) complementing each other to handle different network conditions while transmitting a message that surely overflows the radio buffer (four times the buffer size).
... Here, we focus on the user-generated data-flows complementing the previous investigations in [3], [4] (problem A). These data-flows are sequences of QoS-constrained messages from Command and Control (C2) systems (e.g. ...
... Recent literature, [3], [8]- [15], lacks mathematical arguments as to how such middlewares/brokers will handle everchanging communication scenarios; they usually describe limited experiments using different protocols and radio networks but no general explanation that can be reused to describe/compare any tactical system. We complement the stateof-the-art defining metrics to characterize dataflows which can be reproduced to compare different systems and also to highlight the systems' performance bounds. ...
... within a given time window ω) . Between the users A and the tactical network B there are four functional blocks, (1)(2)(3)(4) in Fig. 1, mapped to an abstract representation for the incoming/outgoing chains i l /o l and the control plane c l with the following functionalities (non-exhaustive list): ...
This paper describes an exploratory study on how to generate sequences of QoS-constrained messages to challenge the underlying store-and-forward mechanisms in tactical networks. The messages come from Command and Control (C2) systems deployed at the tactical edge and the goal is to create reproducible flow of messages with a certain degree of entropy (randomness). Given a mission/operation, we assume that the user-facing services from C2 systems are related to each other and reuse a stochastic model to generate the sequence of messages; here called QoS-constrained dataflows. We studied the system behavior dealing with three different sequences of messages (A1, A2 and A3) to illustrate the computation of metrics using cross-layer contextual information and to highlight the importance of testing tactical systems with different loads. We also compute metrics to characterize the dataflows such as time in the queue, minimum datarate, number of expired messages and so on. Moreover, we used three disruptions patterns in the network to study the sequence of messages being divided in groups so to illustrate and support general conclusions about dataflow characterization. We claim that our methodology can get closer and closer to the performance bounds of store-and-forward mechanisms in tactical networks and can be reproduced by other researchers for quantitative comparisons.
