Distributed Resource Allocation & Congestion Control

Get Complete Project Material File(s) Now! »

V2X communication: History and Overview

The idea of Cooperative Vehicular Networks, initially called Vehicular Ad-hoc Net-work (VANET), has been forked form from the idea of Mobile Ad hoc Networks (MANETs), and with MANETs researchers had an ultimate vision to allow mobile nodes to cooperatively form a network without infrastructure. This has been really challenging without coordination or con guration prior to setup of a MANET in addition to the frequent changing of network topology over a hostile communica-tion medium with nodes having limited power and memory. Thus, out of the many potential application scenarios of MANETs, only VANET has been developed to be deployed in large scale.
In VANETs nodes tend to move in an organized fashion and the direction of movement of nodes is predictable i.e. vehicles follow the direction of the road and can take better advantage of any network infrastructure. Similarly, nodes have better processing power, memory and batteries than MANETs. Moreover, the applications possible through V2X communication can lead to improved road safety, providing value added services and tra c management, thus motivating the industry, researchers and organizations to invest substantially in this domain.

Research History in Europe

In Europe, research work started in the late 1980s, and the rst European project was called PROMETHEUS [12] (1987-1995), which worked on cooperative driving system using vehicular communication at the 57 GHz frequency band. This, was followed by other projects such as, CHAUFFEUR [13] (1996-2000), worked on pla-tooning of trucks. The platoon leader was human driven, and the followers were driven via an electronic tow bar, using V2V communication to transmit deceleration information from the platoon leader to ensure string stability.
Since the beginning of this century, with the advent of GNSS navigation, In-ternet, hardware miniaturization and allocation of a dedicated frequency spectrum for ITS, a lot of Research and Development work, both in the academia and indus-try were initiated in the domain of V2X communication for increasing road safety, driver assistance, tra c management and alike. The German project FleetNet [14] (2000-2003), studied the feasibility of V2X communication based on IEEE 802.11 and UMTS terrestrial radio access (UTRA) to support several types of application with diverse networking requirements, i.e. cooperative driver assistance (safety), oating car data (tra c e ciency), and Internet access (infotainment). Its succes-sor, another German project called Network on Wheels [15] (2004-2008), continued the work of FleetNet. It developed a dual protocol stack for ad-hoc communication between OBUs and OBU and RSU using 802.11p protocol for safety and tra c e – ciency and 802.11a/b/g for communication with the infrastructure for infotainment. These projects contributed to early standardization work at European Telecom-munications Standards Institute (ETSI) and Car2Car Communication Consortium (C2C-CC).
Similar projects followed, such as a project called SAFESPOT [16] between 2006-2010, which worked on V2X communication, safety application, precise rel-ative localization and dynamic tra c map. Another project called GeoNet [17] was conducted between 2008-2010, which worked on combining IPv6 and geonet-working. A project called SEVECOM [18] looked into V2X security and privacy between the years 2006-2009. Another project called COMeSafety [19] was carried out between 2006-2010, which focused on harmonizing and consolidating research results, support standardization and frequency allocation. Most of these projects have contributed to the present V2X standards in Europe.
After the standardization phase, there have been several projects on prototyping and eld operation tests. PRE-DRIVE C2X [20] (2008-2010) performed prototyping and feasibility study of a common European communication system. Its successor DRIVE C2X [21] (2011-2014) tested those on a large scale in seven test sites across Europe.
Finally, another project called Scoop@f [22] (2016-2018), was funded by the European Commission and French government with a goal to deploy 3000 connected vehicles on 2000 km roads across 5 sites in Europe. It performed eld tests of safety applications such as on-board signaling of dangerous events and road hazard warning.

Research History in the USA

Concurrently, in the USA, the academia and the industry have been actively en-gaged in Research and Development on ITS. In 1986, University of California Berke-ley started a research program called California Partners for Advanced Transporta-tion Technology (PATH) 2 to address challenges in California’s surface transporta-tion systems. Currently, it performs research in three domains, i.e. Transportation Safety, Tra c Operations and Modal Applications. Similarly in 1997, United States Department of Transportation (USDOT) started the Intelligent Vehicle Initiative [23], to develop integrated in-vehicle systems with a driver-centric viewpoint.
In 1999 after the availability of a dedicated spectrum of 75 MHz for ITS in the USA, a plethora of research activity took place in the domain of V2X communi-cation. A project called Vehicular Safety Communication (VSC) [24] was carried out between 2002-2004 by 3 organizations i.e. USDOT, Crash Avoidance Metrics Partnership (CAMP) and Vehicle Safety Communications Consortium (VSCC). It de ned the communication needs of safety applications and estimated the feasibility of Dedicated Short Range Communications (DSRC) to satisfy those needs.
Similarly a project called Vehicle Infrastructure Integration (VII) later known as IntelliDrive [25] (2004-2009) looked into V2V and V2I communication for crash avoidance applications and communications. The SafeTrip21 [26] project was car-ried out by the USDOT Research and Innovative Technology Administration (RITA) between 2008 and 2011 for testing and evaluating ITS applications for improving safety, reducing tra c congestion, improving tra c e ciency and transportation convenience. The IntelliDrive project was renamed as Connected Vehicle Research in 2011, and continued research on V2X communication for improving safety, mo-bility and reducing environmental e ects.
The Connected Vehicle Safety Pilot project [27] (2011-2013) was carried out by University of Michigan Transportation Research Institute (UMTRI), CAMP and USDOT ITS program, which performed eld tests to prove the bene t of V2X communication in urban scenarios on real drivers.

Applications & Use Cases of V2X communi-cation

The objectives of V2X communication is to increase road safety, improve trans-portation e ciency, improve ride experience and provide additional services. To this aim, a basic set of applications have been envisioned for Day 1 scenario, which can be categorized as: active road safety, cooperative tra c e ciency and infotain-ment. Tra c e ciency applications provide information for navigation and better route selection, which has not been focused in this thesis. Safety applications intend to improve road safety and have been addressed in this thesis.

Safety Applications for Day 1 Scenario

Safety applications aim to ensure general tra c safety on the road and inform drivers urgently in case of a road hazard or tra c emergency. They function by monitoring the vehicle’s own condition as well as the condition of other vehicles and the road itself, by receiving information via periodic or event triggered messages which complements and extends the range of its on-board sensors.
Safety applications such as Lane Change Warning (LCW) application, help a driver to maneuver carefully, or applications such as Road Hazard Signaling (RHS) increase a driver’s awareness and help the driver to take preventive action and avoid a danger. Similarly, they warn the driver of immediate emergency events, such as hard braking by a vehicle ahead, requiring immediate action to avoid collision. Ex-amples of such applications are: Longitudinal Collision Risk Warning (LCRW), Co-operative Collision Avoidance (CCA), Electronic Emergency Brake Light (EEBL) etc.

Safety Applications for Day 2 Scenario

In future V2X deployment scenarios, the capability of sensors, the computational capacity and the level of autonomy of vehicles will increase, which will enable more advanced V2X applications for more challenging use cases. Highly Automated Driv-ing (HAD) and Platooning are paramount use cases of future cooperative intelligent vehicles. In this regard, vehicles will need to establish a concept called ‘extended horizon’, where vehicles gather information outside the range of their built-in sen-sors, for example a hidden Vulnerable Road User (VRU) around the next building, through cooperative V2X communications and Day 2 safety applications such as Collective Perception. As shown in the example of Figure 1.1, the red vehicle de-tects pedestrians and emits a CP Message (CPM) [28], which alerts the white vehicle before making the right turn.
Similarly, V2X communication capabilities will be used for cooperative driving and navigation, and it is expected that further applications will be developed to exchange a vehicle’s ‘trajectory intent’, i.e. for vehicles to negotiate and coordi-nate their maneuver, using Maneuver Coordination Message (MCM) [29]. Other use cases for future deployment include: vehicle sensor information and state map exchange, cooperative collision avoidance, remote driving, tactile internet, V2X con-nectivity for drones, bird’s eye view via drones etc.
These Day 2 applications will transmit a variety of messages in the channel, which will lead to channel congestion. In this thesis, we analyze that existing channel congestion control protocol standardized for Day 1, a.k.a Decentralized Congestion Control (DCC) in European standards, use the channel ine ciently. The main contribution of this thesis is to improve those congestion control protocols and propose better ones for managing multiple V2X safety messages in a congested channel

Transmission Technology & Spectrum

As mentioned earlier, two leading wireless communication technologies have been developed for V2X communication: IEEE 802.11p based ITS-G5/DSRC and 3GPP LTE V2X.

IEEE 802.11p based ITS-G5/DSRC

ITS-G5/DSRC operates at the 5.9 GHz band, using IEEE 802.11p PHY and MAC layer protocol, which has been derived from IEEE 802.11a with modi cations to adapt it to the dynamic context of V2X communication. At the PHY layer it uses Orthogonal Frequency Division Modulation (OFDM), with 52 subcarriers (48 data and 4 pilot subcarriers) which are placed within 10 MHz wide channels. Compared to 802.11a, the subcarrier spacing is halved (156.25 kHz instead of 312.5 kHz), which doubles the time domain parameters, to cope with Inter-Symbol Interference due to Multi-path fading in challenging propagation scenario.
The Medium Access of IEEE 802.11p uses the legacy IEEE 802.11 Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA), listen before talk approach, having a di erent MAC functionality per channel. Similarly, for Quality of Service (QoS) 802.11 uses Enhanced Distributed Channel Access (EDCA) which originates from IEEE 802.11e-2005, o ering 4 channel access priorities, i.e. Voice, Video, Best E ort and Background. Lastly, one key di erence of 802.11p compared to infras-tructure Wi-Fi is that it operates in an ad-hoc mode called Outside the context of a BSS (OCB). In this mode, nodes do not form a basic service set (BSS), and communicate on the y to avoid the delay caused by network setup steps like chan-nel scanning, authentication, and association. ITS-G5 will be detailed further in Chapter 2 of this thesis, along with the other Access technology, i.e. 3GPP LTE V2X.

Channels & Frequency spectrum

Since 2008, three 10 MHz channels have been allocated in Europe, for safety related applications at the 5.9 GHz band (called ITS-G5A at ETSI), with one CCH between 5895 and 5905 MHz and two SCHs between 5875 and 5895 MHz, as shown in Figure 1.2. Similarly, four additional channels have been reserved for ITS communication, called ITS-G5B (5855-5875 MHz) and ITS-G5D (5905-5925 MHz). Altogether, this makes a 70 MHz bandwidth between 5855 MHz and 5925 MHz reserved for ITS applications in Europe. Another ITS band called ITS-G5C actually corresponds to eleven 20 MHz channels of the lower RLAN band currently allocated for WiFi communication between 5.5 GHz to 5.72 GHz.
Although reserved since 1999 in the USA and 2008 in Europe, Day 1 connected and cooperative vehicle applications merely use one of these seven channels, the six others being planed for Day 2.

Distributed Resource Allocation & Conges-tion Control

In IEEE 802.11 based vehicular networks, there is no centralized channel resource allocator and the nodes need to prevent channel saturation by periodically moni-toring the channel load, measured via Channel Busy Ratio (CBR), and limiting the spatial channel usage via Transmit Power Control (TPC) and/or temporal channel usage via Transmit Rate Control (TRC). In ETSI congestion control a.k.a DCC standards, TRC has been speci ed as the principle mechanism for congestion con-trol at the Access layer [5].
DCC Access limits the transmit rate using tra c shaping at the Access Layer, via queuing and ow control above the EDCA queues, as a function of the CBR, as shown in Figure 1.3a. Packets from applications are enqueued in one of the four queues and are prioritized using Tra c Class (TC). Flow control is managed via a single leaky bucket for all the queues, which allocates transmit opportunity to a lower priority queue only if a higher priority queue has no packet. Figure 1.3b shows a zoomed in view of the queues with di erent V2X safety messages. CPM and MCM have been marked in grey as their exact priorities have not been speci ed in the standards.
The rate of the leaky bucket depends on the CBR, i.e. for higher CBR it allows a lower transmit rate and vice-versa. Similarly, the mapping of CBR to leaky bucket transmit rate can be of two variants, i.e. Reactive and Adaptive DCC. Reactive DCC maps the transmit rate to CBR using a state machine, while Adaptive DCC iteratively adapts the transmit rate to reach a target CBR. However, for the highest priority TC, the leaky bucket can be bypassed via a token bucket. DCC is being extended to be a cross-layer mechanism, with functionalities at the Facilities [30] and Network [31] layers and a cross-layer Management entity [32], which have been detailed in Chapters 2 and 3 of this thesis.
By controlling the transmit rate of each node and of the services of a node, DCC can be seen as to allocate transmit opportunity or communication resources for each node and its services. Therefore, channel resource is allocated not only via CSMA/CA, but DCC additionally controls the resource allocation to each node on top of CSMA.

READ  Optimal scheduling of household appliances with battery storage and coordination 

Research Problem

V2X channel congestion control has been investigated and standardized for initial V2X communication scenario, considering mostly a single type of safety message on the control channel. However, as discussed in Section 1.3, in future scenarios vehicles will broadcast a variety of messages, which will easily lead to channel congestion. This raises the following research problems which have been addressed in this thesis.
1. ETSI DCC uses the channel capacity ine ciently as found in several previ-ous studies for a single message [33, 34], and multiple messages [35, 36]. The TRC has been found to be over aggressive, with an unstable control pro-cess. Although DCC has been revised recently [5], but some of the original problems with rate control with Reactive DCC still remain. The rst re-search problem is how to improve the performance and stability of Reactive DCC transmit rate control for multiple applications.
2. There are 4 DCC queues for tra c shaping and ow control, as shown in Figure 1.3a. In future, a vehicle will transmit more than 4 types of messages. Therefore, messages from di erent applications will have to share the same queue, as shown in Figure 1.3b. However, at the Access layer there is no notion of application but only TC, so QoS cannot be enforced per application. Similarly, packets from one application may erase packets of other applications waiting in the queue. The second question is, how to provide QoS per application, instead of per TC and how to orchestrate channel resources among applications with homogeneous and heterogeneous priority.
3. Prioritizing applications using static priority will inde nitely starve low pri-ority applications during channel congestion. The third problem is, how to dynamically characterize the priority of safety V2X applications, providing a node more control and exibility for resource orches-tration among multiple applications, instead of using static TC.
4. In Day 2 scenario vehicles will have non-homogeneous capabilities and com-munication requirements. However, DCC allocates similar channel access op-portunity to neighboring nodes facing similar channel load. The fourth re-search problem is, how to decentrally distribute channel resources asymmetrically among vehicles with di erent needs and number of applications.

Research Methodology and Design Choices

In this section we describe our approach to address and solve the aforementioned problems, along with the reasoning behind the design choices.
Step 1: Performance Improvement of DCC Rate Control
Firstly, we analyzed the performance of Access DCC for controlling transmit rate of multiple applications, and Reactive DCC uses ine cient rate control parame-ters along with non-optimal adaptation to channel load, resulting in an unstable control process. We propose less severe rate control parameters, with continuous and smooth rate adaptation using a memory, instead of abrupt rate oscillation of standardized DCC. This largely improves the transmit rate of multiple applications, controlled by Reactive DCC while limiting the channel load.
Step 2: Evaluation of DCC Tra c Shaping
Afterwards, we evaluated the impact of DCC queuing and tra c shaping. Several queuing methods were tested, and performance issues were found with those. In-stead of queuing at the upper MAC layer, we concluded that the intelligence of traf-c shaping should be shifted from the Access to the Service/Facilities Layer, which has higher intelligence and more exibility to orchestrate channel access among the applications.
Step 3: Connecting Access DCC with the Service Layer
Reactive DCC limits transmit rate w.r.t CBR, without considering the packet du-ration, causing ine ciency as channel resource is a product of the transmit rate and duration. Although, the recently revised DCC standard [5] proposes rate limits for two packet duration ranges, but a lack of continuous relation of the packet duration, makes the Reactive Access DCC incompatible for managing multiple applications. Therefore, we included a third dimension, i.e. continuous function of the packet size to Reactive Access DCC, to make it compatible for handling multiple applications.
Similarly, spreading the DCC functionality at di erent layers, requires well co-ordination among those entities. ETSI has proposed cross-layer DCC mechanisms [32], but we observed issues in those protocols, resulting in missed transmit oppor-tunities. We propose a resource orchestrator (discussed in Step 5) at the Facilities layer to mitigate the problem.
Step 4: Flexible & Dynamic Characterization of Applications
Using static QoS classes inde nitely starves low priority applications during chan-nel capacity shortage. Therefore, we propose exible characterization of V2X safety applications, to give each node more control for allocating channel resource among the applications. Similarly, the characterization depends on contextual factors, such as rank, usefulness and urgency, where no application can inde nitely be starved of transmit opportunity.
Step 5: Implementing a Service Layer Application Resource Orchestra-tor
Building on the previous steps, we propose a design to shift the tra c shaping intel-ligence, from the Access to the Service layer using a centralized in-vehicle resource orchestrator in the protocol stack, while leaving only the task of ow control at the Access layer.
The orchestrator characterizes applications according to Step 4, and allocates transmit opportunity among multiple applications using a budgetary scheduler based on resource earning/spending supporting a smooth resource allocation over time. It allows exible adjustments in time of the priority between V2X services as a function of their dynamic budget.
Step 6: Resource Allocation for vehicles with Heterogeneous Needs After dealing with in-vehicle resource allocation, we move to inter-vehicle resource allocation addressing the heterogeneous needs of each vehicle. We propose a coop-erative and distributed mechanism for vehicles to identify the communication needs and importance of other vehicles sharing the channel, and sacri ce a proportional amount of resources from their own lower priority messages, which get used by vehicles with higher priority demands, while maintaining the overall channel load below the saturation level.
Step 7: Machine Learning for Predicting Resource Availability
Lastly, we also investigated using machine learning for predicting resource availabil-ity and application transmission patterns, in order to better allocate and reserve transmit opportunities for applications. Nevertheless, as the work is not mature enough to be integrated with the rest of this thesis, it has been put in the Ap-pendix.

Contribution

The key contributions of this thesis can be summarized as follows:
1. We design an application resource orchestrator at the service layer.
2. We highlight the inadequacy of static tra c class for classifying safety V2X applications and propose dynamic prioritization using contextual parameters.
3. We demonstrate ine ciency of Reactive Access DCC, and improve its perfor-mance
4. We demonstrate issues of the ETSI DCC architecture, mainly Access layer queuing and ow control and highlight the necessity of a DCC entity at the Facilities Layer.
5. We demonstrate the challenges of cross-layer coordination of ETSI DCC, and propose a more compatible design.
6. We propose decentralized resource allocation and congestion control for vehi-cles with diverse capabilities and applications having heterogeneous channel resource requirement.
7. We demonstrate the challenges of ITS-G5 spectrum sharing with Wi-Fi at the 5.9 GHz band and evaluate the e ects on the performance of V2X safety communication.
The applicability of this work is not limited to DSRC/ITS-G5, and the implica-tions are also valid for channel congestion control for LTE V2X Mode 4. Considering that applications and tra c pattern remain the same, congestion control in LTE V2X may also face similar cross-layer coordination issues which have been analyzed with ITS-G5 in this thesis. Consequently, with further work, the contributions of this thesis can be extended to LTE V2X, as discussed in Chapter 5.

List of Publications

1. In vehicle resource orchestration for multi-V2X services Khan, Mohammad Irfan; Sesia, Stefania;Harri, Jer^ome VTC 2019 Fall, 90th IEEE Vehicular Technology Conference, 22-25 September 2019, Honolulu, Hawaii, USA
2. Deep learning-aided resource orchestration for vehicular safety communication Irfan Khan, Mohammad; Aubet, Francois-Xavier; Pahl, Marc-Oliver; Harri, Jer^ome Wireless Days 2019, IEEE/IFIP Days 2019, 11th edition, 24-26 April 2019, Manchester, UK
3. Flexible packet generation control for multi-application V2V communication Khan, Irfan; Harri, Jer^ome VTC 2018-Fall, IEEE 88th Vehicular Technology Conference, 27-30 August 2018, Chicago, USA
4. Integration challenges of facilities-layer DCC for heterogeneous V2X services Khan, Mohammad Irfan; Harri, Jer^ome IV 2018, 29th IEEE Intelligent Vehi-cles Symposium, 26-29 June 2018, Changshu, Suzhou, China
5. IoT and microservices based testbed for connected car services Datta, Soumya Kanti; Khan, Mohammad Irfan; Codeca, Lara; Denisy, B.; Harri, Jer^ome; Bonnet, Christian SMARTVEHICLES 2018, 5th WOWMOM Workshop on Smart Vehicles: Connectivity Technologies and ITS Applications, colocated with WOWMOM 2018, June 12-15, Chania, Greece
6. Rethinking cooperative awareness for future V2X safety-critical applications Khan, Mohammad Irfan; Hoang, Gia-Minh; Harri, Jer^ome VNC 2017, IEEE Vehicular Networking Conference, November 27-29, 2017, Torino, Italy.

Table of contents :

1 Introduction 
1.1 Introduction
1.2 V2X communication: History and Overview
1.2.1 Research History in Europe
1.2.2 Research History in the USA
1.3 Applications & Use Cases of V2X communication
1.3.1 Safety Applications for Day 1 Scenario
1.3.2 Safety Applications for Day 2 Scenario
1.4 Transmission Technology & Spectrum
1.4.1 IEEE 802.11p based ITS-G5/DSRC
1.4.2 Channels & Frequency spectrum
1.5 Distributed Resource Allocation & Congestion Control
1.6 Research Problem
1.7 Research Methodology and Design Choices
1.8 Contribution
1.9 List of Publications
1.10 Organization of the thesis
2 State of the Art 
2.1 Standardization Organizations
2.2 Standardized Protocol Stack in Europe and USA
2.2.1 Architecture – ETSI TC ITS
2.2.2 Architecture – IEEE 1609.4 WAVE
2.2.3 ETSI Facilities Layer & V2X Safety Messages
2.2.4 SAE Message Sub Layer & Standardized messages in the USA
2.2.5 ETSI Network & Transport Layer
2.2.6 IEEE Wave Network & Transport Layer
2.2.7 ETSI Access layer
2.2.8 IEEE Wave Access layer
2.2.9 3GPP LTE V2X
2.2.10 Cross Layer Entities
2.3 Channel Coexistence
2.4 Channel Congestion Control
2.4.1 ETSI Decentralized Congestion Control Standards
2.4.2 SAE Congestion Control Standard
2.5 Dierences between standards in Europe and USA
2.6 Literature Review: Channel Congestion Control
2.6.1 Transmit Rate Control
2.6.2 Transmit Power Control
2.6.3 Combined Transmit Rate and Power Control
2.6.4 Controlling other parameters
2.6.5 Awareness Control
2.6.6 Channel Congestion Control for Multiple Packet Types
2.6.7 Beyond State of the Art
2.6.8 5.9 GHz Spectrum Sharing between ITS-G5/DSRC and other technologies
2.7 Existing Design Philosophy
2.7.1 Bottom Up Approach
2.7.2 Our Proposition: Top Down Approach
3 Problem Statement
3.1 Introduction
3.2 General Characteristics & Challenges of Vehicular Networks
3.2.1 Decentralization
3.2.2 Dynamicity
3.2.3 Heterogeneous Network Trac Pattern
3.2.4 Heterogeneity in Communication Resource Requirement
3.2.5 Losing the dedicated spectrum for ITS communication
3.3 System Description & Key functions of Congestion Control
3.3.1 Inter{Vehicle Resource Allocation
3.4 In-Vehicle Transmit Rate Control
3.4.1 Classifying Services for limited channel resource Attribution
3.4.2 Access Layer Rate Control
3.4.3 Service/Facilities Layer Rate Control
3.5 Challenges of Inter Vehicle Transmit Rate Control
3.5.1 Asymmetric Resource demand by heterogeneous Nodes
3.5.2 Distributed Asymmetric Resource Allocation
3.6 Challenges of In Vehicle Transmit Rate Control
3.6.1 Limitations of Priority Queuing:
3.6.2 Limitations of Trac Class and Access Category
3.6.3 Limitations of Access DCC Queuing Policy & Queue Size
3.6.4 Limitations of Channel Load Measurement
3.6.5 Limitations of Reactive Transmit Rate Control
3.6.6 Challenges of Cross-Layer Dependency
3.7 Paradigm Shift: Innovation and Methodology
3.7.1 Need for an In-Vehicle Resource Orchestrator
4 Results and Analysis
4.1 Introduction
4.2 Simulator & Simulation Scenario
4.2.1 iTetris-NS3 Network Simulator
4.2.2 Simulation Parameters
4.3 Access Layer Transmit Rate Control
4.3.1 Performance Benchmark: No Congestion Control
4.3.2 Reactive DCC – Earlier Version
4.3.3 Reactive DCC – Latest Version
4.3.4 Improving Reactive DCC
4.3.5 Adaptive DCC
4.4 DCC Access Queuing
4.4.1 Elect of Trac Class & Packet Generation rate per Service
4.4.2 DCC Queuing: Same Trac Class & Equal Packet Generation per Service
4.4.3 Elect of Queue Size
4.4.4 DCC Queuing Delay
4.5 DCC Cross Layer Coordination
4.5.1 Packet Generation Control via Transmit Rate Limit on top of Reactive Access DCC
4.5.2 Packet Generation Control via Transmit Rate Limit on top of Adaptive Access DCC
4.5.3 Packet Generation Control via Channel Resource Limit on top of Adaptive DCC
4.5.4 Facilities DCC on top of Improved Reactive DCC
4.6 Facilities Layer Resource Orchestration
4.6.1 Application characterization function
4.6.2 Application Resource Calculation
4.6.3 Packet Scheduling by the Resource Orchestrator
4.6.4 Resource Orchestration Performance Evaluation
4.7 Inter-Vehicle Resource Allocation (Asymmetric resource demand by vehicles)
4.7.1 Example of Distributed Resource Re-allocation
4.7.2 Resource Re-allocation Mechanism
4.7.3 Evaluation with Static Scenario
4.7.4 Evaluation with Dynamic Scenario
4.8 ITS-G5 & WiFi Spectrum Sharing
4.8.1 Performance Evaluation
4.8.2 Articial Scenario: 3 Zones of Awareness
4.8.3 Scenario: Outdoor WiFi
4.8.4 Scenario: Indoor WiFi
5 Conclusions and Discussion
5.1 Summary of Problems Addressed and Contributions
5.1.1 DCC Access
5.1.2 DCC Queuing and Trac Shaping
5.1.3 Design and Cross Layer Issues
5.1.4 Resource Orchestrator
5.1.5 Inter-Vehicle Asymmetric Resource Allocation
5.1.6 Spectrum Sharing
5.2 Future Work
5.2.1 In-vehicle resource allocation
5.2.2 Dynamic prioritization for Inter-vehicle resource allocation
5.2.3 Network Slicing for DSRC/ITS-G5
5.2.4 Machine learning for resource allocation

GET THE COMPLETE PROJECT

Related Posts