Distributed Edge Computing to Assist Ultra-Low-Latency VANET Applications

Vehicle ad hoc networks (VANETs) are a class of peer-to-peer wireless networks that are used to organize communication between cars (V2V), between cars and infrastructure (V2I), and between cars and other types of nodes (V2X). These networks are based on the dedicated short-range communication (DSRC) 802.11 standards and are mainly intended for organizing the exchange of various types of messages, mainly emergency ones, to prevent road accidents, alert when a road accident occurs, or control the priority of the roadway. Initially, it was assumed that cars would only interact with each other, but later, with the advent of the concept of the Internet of things (IoT), researchers began to analyze connectivity with other devices. This, in general, will allow the combination of various road users with other devices that can used in the creation of intelligent transport infrastructure in a single smart city management system. Infrastructure is necessary for the provision of services, monitoring, and management of the VANET. As infrastructure objects, stationary roadside units (RSUs) have been proposed. The aim of this paper is to analyze the use of mobile edge computing to decrease the load to the base station and latency between RSU clouds and to provide a real experiment using software-defined networking and mobile edge computing for RSUs.


Introduction
Based on the pre-planned road map announced by the International Telecommunication Union (ITU) and the Third Generation Partnership Project (3GPP), by 2020 it is expected to begin a new era of mobile communication system with great and efficient capabilities with the announcement of the fifth generation of mobile communication systems (5G) [1]. By the end of this year, standards for building the 5G cellular system with the announced requirements will be ready and the final step in the 5G road which is the system implementation will begin. With this great achievement, we will have a new era of telecommunication systems represented by the 5G system and its use cases. Each use case represents a challenge and required certain design aspects to be real [2]. Most of these use cases will be available and can be implemented with the realization of the 5G system except the Intelligent transportation system like Vehicle ad hoc network (VANET).
The 3GPP identifies five main categories of 5G use cases, with potential services for each category [3].
These categories can be reduced and distributed over a three major verticals of 5G services. These verticals are also defined by ITU-R as the three usage scenarios for 5G and beyond, which The system employs cloud units at the edge of the radio access network (RAN) based on the concept of Mobile Edge Computing (MEC). MEC is a new trend established by the cellular network operators to improve the whole network efficiency by offloading its operations to nearby clouds. European Telecommunications Standards Institute (ETSI) is one of the main organizations concerned with the MEC [8]. ETSI announced an Industry Specification Group (ISG) known as MEC to research and standardize the new technology. Simply MEC can be defined as the way of moving cloud computing capabilities to the edge of the mobile networks.
Moving from the great, massive and expensive data centers into small distributed cloud units based on a small hardware platform will open the way for achieving the required latency constraint for tactile realization. Moving cloud units only one or two communication hops away from the end user is a key solution for the 1ms end-to-end latency. Moving cloud computing to the edge of the mobile produces a lot of benefits that can be summarized in the following points [9]: 1-reduces the round trip latency of communicated data, 2-provides an efficient way for offloading data delivered to the core network, 3-provides high bandwidth, 4-introduces new services and applications by accessing the network context information, and 5-introduces new services and applications by accessing the network context information.
Our SDN lab propose an intelligent core network for tactile internet in [10], that can be employed side by side with the concept of multi-level edge computing to provide an appropriate structure for VANET systems [11].
The next step concerned with the edge computing units is related to simulate the remote environment on the edge cloud unit and expect its behavior by means of AI. This represents a critical part toward system realization to achieve the required latency.

Related works
Currently, research in the field of alternative methods of transport network management is gaining momentum.
A group of researchers from University of Toulouse reviewed the SDN hybrid architecture for ITS managing [12]. The key idea of this architecture is the hybrid control level, which includes the base station controller of the mobile network and the controller of the RSU modules, as well as the central controller to coordinate actions of different controllers. The central controller creates a global view of the network infrastructure by sending information to controller of each network with data in the cloud. It sends each controller global rules describing the general behavior of the network, while BS and RSU controllers define specific rules that must be set in each network device. Communication between SDN controllers is performed using a special interface, known as East-West, also communication between SDN controllers and the cloud is performed through API interfaces.
The aim of the experiment was to demonstrate how a global view of the network, combined and enriched with information obtained from the cloud, allows to more effectively manage network behavior by providing ultimately a service with increased performance.
Another research project on this subject is the work group from the University of Singapore (School of Electrical and Electronic Engineering Nanyang Technological University, Singapore) [13]. This approach also involves the use of existing mobile network architecture, in addition to the implementation of management based on SDN technology. It should be noted that all the VANET network architecture is divided into segments (Control Region), which may contain a certain amount of RSU, which operates a control member (Local Controller), this significantly reduces the delay in case of transmission of information to the network management unit.
The central network controller is located in its core. The interaction between the central controller and local controllers takes place based on mobile communication network. Authors called this architecture SDVN -Software-Defined Vehicular Network.
The key parameter of simulation and mathematical models, which was chosen in this paper is delay. For comparison, three network architectures were selected: VANET network based on the 4 of 16 AODV routing protocol, SDVN with a central controller and SDVN with controllers located within the segment.
In [14], authors proposed a context-aware packet-forwarding mechanism for ICN-based VANETs. The relative geographical position of vehicles, the density and relative distribution of vehicles, and the priority of content were considered during the packet forwarding.
In [15] authors focused on a vehicle-to-vehicle communication system operating at a road intersection, where the communication links can be either line-of-sight (LOS) or non-line-of-sight (NLOS). Authors presented a semi-empirical analysis of the packet delivery ratio of dedicated short-range communication (DSRC) safety messages for both LOS and NLOS scenarios using a commercial transceiver.
Authors in [16] purposed critically review the existing methods of adaptive traffic signal control in a connected vehicle environment and to compare the advantages or disadvantages of those methods The proposed architecture provides load balancing between distributed controllers and uses a local view of the network to make better decisions. Both theoretical model and simulation results confirm that the proposed SDVN architecture differs from the existing SDVN with a central controller in terms of latency. The proposed system has the same low latency as the existing VANET network.

System structure
The IEEE 802.11 standard provides two modes of operation -infrastructure (BSS, ESS) and non-infrastructure (IBSS). The last one is a new class of networks, called Ad hoc or target networks. Now within the framework of Ad hoc networks classification are divided into mobile and fixed networks. Mobile ad hoc networks (MANET -Mobile Ad hoc Network) [17] include Ad hoc networks for vehicles a standard for that was developed as part of the IEEE 802.11p working group [18][19]. For the interaction between vehicles and with the public use network in the coming years can lead to the formation of a new, very large-scale segment of the telecommunications market. Already a modern car integrates a GPS / GLONASS receiver, various sensors, an on-board computer. However, the task that is posed when creating a VANET is somehow different [20]. First of all, create a network interface in the car that would allow supporting four groups of connections: car -car, car -infrastructure network. Technical means of the IEEE 802.11p standard should operate at speeds up to 200 km / h and at distances up to 1 km. The physical layer and the MAC sublayer are based on the IEEE 802.11a standard. The frequency range for the USA includes the spectrum from 5.859 to 5.925 GHz, for Europe, the Electricity and Postal Services Commission (CEPT) recommends the use of two subbands 10 MHz wide each: 5.865 -5.875 GHz and 5.885 -5.895 GHz. Architecture studied VANET network without the use of SDN is based on a centralized remote management of RSU via base station controllers, which in turn are connected to the main controller. Controllers transfer data to the cloud server. At the same time, the RSU service control packages are located in the same network with the traffic that comes from the On-Board Unit (OBU).

VANET network Challenges
Traditional VANET networks are characterized by maximum decentralization, when there is no dedicated server in the network, and the entire infrastructure is distributed to communication centers. This feature brings following disadvantages: • low mobilization abilities; • long system response time to external influences. To date, VANETs support a large number of new services and protocols. Nevertheless, there are still a number of challenges that impede the full implementation of this technology in real life. a) Problem ensuring noise immunity For transport networks VANET is a large amount of interferences of different types. The effect of interference can be maximized by choosing the optimal channel frequency or by adjusting the power of the receivers and transmitters of the devices in the network. However, in classical VANET networks, this problem remains unresolved and has a significant impact on the network operation. b) The problem of ensuring the security of transmitted data Causes of problems related to the information security of VANET transport networks are: • lack of means to protect nodes from intruders and hackers; • the ability to listen to the channels and the substitution of messages due to the general availability of the transmission medium; • inability of applying the classical security system due to the peculiarities of the classical architecture of the VANET networks; • the need to use complex routing algorithms that take into account the likelihood of incorrect information from compromised nodes as a result of changes in the network topology; • Any node that is in the signal source range that knows the transmission frequency and other physical parameters (modulation, coding algorithm) can potentially intercept and decode the signal. At the same time, neither the signal source nor the recipient will know about it; • inability to implement security policy due to the peculiarities of the classical architecture of VANET networks, such as the absence of a fixed topology and central nodes.
c) The problem of routing efficiency In VANET networks, the bandwidth problem is most acute with the simultaneous transmission of a large amount of video information. The situation get worsen by the inefficiency of the routing methods, when the network becomes overwhelmed with broadcast requests or a bottleneck is formed. An example of inefficient routing is shown 2 in the figure when all traffic transit through one node. In this paper we considered possibilities of solving some problems that occur when using the local SDN controller, the OpenFlow protocol can be used as a control over radio channels. OpenFlow SDN with minimal latency can safely solve the RSU control problem. The implementation of SDN MEC technology will partially solve the security problem. The concept of SDN implies the use of separate control channels and traffic subscriber. Thus, in DDoS case attack from the OBU or the alleged OBU, the RSU control channel will not suffer. And the SDN controller can take any measures to solve the problem (temporarily disable the RSU, tune the RSU to other radio channels, for example).

Traffic model
Approximate value of waiting time in the packets queue in RSU, with linear packet generation. G/G/1 queuing model: In this case = 0, because the data rate on all channels between the cars is the same, the variance of service time equals to 0.
The service time is determined by the packet length and data rate: The ratio of the generated traffic to the data rate in the channel: Package delivery time: Coefficient of variation of the generated traffic: In view of the above, the average waiting time can be estimated by the formula: approximate average speed in RSU and MEC

Offloading algorithm based on edge computing
This structure is based on the use of software-defined network SDN and offloading the core network by unloading traffic to the border of the radio access network (RAN), namely, MEC technology of mobile edge computing. The RSU offload traffic to the edge of the RAN network, after which the traffic is directed to the SDN controller and the cloud server.

Experimental work and results
Analysis of the results of full-scale experiment to identify key indicators for the VANET network was based on traffic dump recorded on the radio interface of one of the devices. The traffic was recorded using the utility built into the device. The experiment was performed in open space and the maximum removal of network elements from each other was 300 m. Based on the tests performed, the following set of WSMP protocol package parameters was chosen, which will be enough for statistics:  The total packet size is 255 bytes. The data presented above were obtained from the test results of a segment of the transport self-organizing network will and be used in the future for modeling and calculating the RSU load. The key parameters of the functioning segment of VANET network model are presented in the Table1. To determine the effectiveness of the VANET network with control based on software-defined networking, a serie of 5 experiments was conducted for each of the architectures shown on Figure 4. The duration of each experiment was 15 hours. During the experiment, 30 Raspberry pi devices were used to imitated OBU vehicles, the number of RSUs was 10 devices. The study scenario implied a serial connection of the OBU to the RSU at specified intervals, which was used to simulate the movement of vehicles. In situations where more than three OBUs were connected to one RSU, an additional delay of 5 ms was introduced. The results of the experiments are shown in Figure 4. According to the experimental results the average packet loss amounted to 0.082%, thus it is possible to conclude that the controller is coped with the goal of reducing the network load, thus confirming the efficacy of the test approach.
1) The following proposed model was tested for the dependency of the packet delivery time on the distance between RSU stations.
A. Three cases were considered, with a flow rate of 2000 cars per hour and speeds of 30 km / h, 50 km / h and 60 km / h, respectively, the size of the packets was 30 bytes. Based on the data obtained, the best packet delivery time is observed that at a speed of 60 km / h. Moreover, with increasing distance between the RSUs, the packet delivery time increases.
B. Next, we consider a model with a flow rate of 3000 cars per hour:  The shortest packet delivery time for cars moving at 60 km / h and twice the delivery time for packages for cars moving at 30 km / h.

System Realization
During the experiment were realized RSU, which are included with the traffic lights on the road. The use of these devices in the aggregate will give the possibility of unloading road traffic and allow emergency services to travel without traffic jams.  Table 2 shows the characteristics of the board RouterBOARD 435G   Next, for experimental needs, was select the scale of the built pilot zone in Saint Petersburg (Russia): 1) The section of "Bukharestskaya Street" between intersections with "Salova Street" and "Prospekt Slavy".
2) The section of Salova Street between intersections with "Alexander Kovanko Street" and "Bukharestskaya Street".
3) The section of "Aleksandra Kovanko Street" from the entrance of the Bus depot No. 1 of the Saint-Petersburg State Unitary Enterprise "Passazhiravtotrans" to the intersection with "Salova Street". Type of modulated pilot zones shown on the map figure:

Conclusions
Currently, the situation on the roads is becoming increasingly tense, regarding to the increasing of the number of cars, the scaling of the road transport network, population growth and many other factors. To optimize and automate traffic, increase driver and pedestrian safety, monitor traffic violations and intelligent transport systems, the load on which is constantly increasing. To solve problems with ITS, it is necessary to organize a competent management system. Current solution is to use the concept of software-defined networking for Intelligent transport systems management. SDN can reduce workload, optimize network infrastructure architecture that simplifies scaling of the network, improve the security of transmit information. During the work on this subject have been considered the most promising approaches to the management of ITS, identified key indicators of the functioning of VANET network segment, experimentally proved the advantages of software-defined network management for ITS. A series of experiments were conducted with different densities, different speeds at different locations autoflow RSU along the road. The results were analyzed and it was composed RSU service model appropriate model QS G / D / 1. The results of this work allow us to consider the concept of SDN and MEC for ITS management, as well as to continue research in this direction, given the trends in the development of modern communication networks.