华为Use+Cases%2C+Requirements%2C+and+Design+Considerations+for+5G+V2X+.pdf

1Use Cases,Requirements,and Design Considerationsfor 5G V2XMate Boban,Apostolos Kousaridas,Konstantinos Manolakis,Joseph Eichinger,Wen XuHuawei Technologies,German Research Center,80992 Munich,GermanyEmail:AbstractUltimate goal of next generation Vehicle-to-everything(V2X)communication systems is enabling accident-free cooperativeautomated driving that uses the available roadway efficiently.Toachieve this goal,the communication system will need to enable adiverse set of use cases,each with a specific set of requirements.Wediscuss the main use case categories,analyze their requirements,and compare them against the capabilities of currently availablecommunication technologies.Based on the analysis,we identify a gapand point out towards possible system design for 5G V2X that couldclose the gap.Furthermore,we discuss an architecture of the 5GV2X radio access network that incorporates diverse communicationtechnologies,including current and cellular systems in centimeterwave and millimeter wave,IEEE 802.11p and vehicular visiblelight communications.Finally,we discuss the role of future 5GV2X systems in enabling more efficient vehicular transportation:from improved traffic flow through reduced inter-vehicle spacing onhighways and coordinated intersections in cities(the cheapest wayto increasing the road capacity),to automated smart parking(nomore visits to the parking!),ultimately enabling seamless end-to-endpersonal mobility.I.INTRODUCTIONPersonal mobility and vehicular transportation systems in gen-eral are undergoing somewhat of a revolution.The reasons forthis can be found in the new societal and market trends.The mainnew societal trends affecting the transportation are:i)new waveof urbanization creating pressure on the existing transportationinfrastructure,which cannot grow as fast as the demand;ii)evermore stringent emission-and energy-related regulation;and iii)high pressure on public transport and logistics/delivery servicesto become more adaptive and dynamic.The key market trendsare:i)the advent of automated driving;ii)new modes of car useand ownership(i.e.,a shift towards the“shared economy”);andiii)live and open data availability,including crowd sourcing andopen platforms,which enables more efficient use of transportationresources.These trends are creating a shift towards more reactiveand intelligent transport infrastructure,with the following goals:Accident-free transportation;Supporting higher traffic flow(i.e.,increasing the road sur-face utilization n-fold,currently standing below 10%1);Higher vehicle utilization(e.g.,increasing the average per-sonal car utilization well above the current 5%);More efficient/greener transport(zero emission vehicles).Communication technologies,in the form of Vehicle-to-everything(V2X)communication1,will play a key role in reach-1In terms of the naming conventions,we consider that the term Vehicle-to-everything(V2X)subsumes all of the following communication modes:a)Vehicle-to-Vehicle(V2V);b)Vehicle-to-Infrastructure(V2I)(e.g.,communicationwith roadside units(RSUs),traffic lights,or,in case of cellular network,basestation);c)Vehicle-to-Pedestrian(V2P);and d)Vehicle-to-Network(V2N),wherethe vehicle connects to an entity in the network(e.g.,a backend server or a trafficinformation system).ing these goals.While the in-vehicle sensors can enable manyfunctionalities without the use of inter-vehicle communication,thebenefits that V2X communication promises are:i)safer driving asa consequence of enabling the well-studied safety use cases 2,3;and ii)improved road capacity due to better road 4 andparking infrastructure 5 utilization.The goals of this paper areto:consolidate and analyze relevant use cases and their require-ments that will drive the 5G V2X communication systemdesign;discuss the capabilities of existing communication technolo-gies to support the 5G V2X use cases,including the resultinggap analysis;provide guidelines on how existing and future communica-tion systems can be leveraged as part of a 5G V2X solution.As a starting point,below we elaborate on the potential benefits ofV2X communication through an illustrative example that involvesmany of the use cases described in 2 and 3.Figure 1 shows an aspirational example of what future personalmobility might look like,exemplified through several salient usecases,such as those found in 2 and3,some of whichare shown in Fig.2.The following use cases are employed inconsecutive steps and denoted in Fig.1 accordingly).1)Vehicle-on-demand:a vehicle user2hails a car via an appdepending on purpose(e.g.,on a weekday:family sedan;on a sunny weekend:convertible).2)The vehicle self-drives or is tele-operated 6(either byhuman or remote robot/artificial intelligence)to the user.Users personal transportation app connected over the mo-bile radio network adjusts the car to his presets.3)The vehicle is capable to drive itself:it searches for theplatoon best suited for the vehicle users preferences andperforms cooperative maneuvering to join selected platoon,thus saving energy by allowing car following at very shortgaps.4)Two-way navigation(i.e.,feeding and obtaining real timetraffic info)guides the vehicle around a congested stretchof the road.5)The user takes over control to exit the highway and driveby himself on a particularly scenic road suggested by theconnected navigation system.6)The vehicle drops the user off at the destination.7)The vehicle self-drives 7,either to an automated parkinglot 5 7a)or to a different user 7b).Automatedelectric vehicles might not need parking garages in the2Note that we use the term“vehicle user”to distinguish the new mobilitymodels from conventional“driver”and“passenger”models.arXiv:1712.01754v1 cs.NI 5 Dec 20172Fig.1.Motivating example for 5G V2X.To enable the described scenario,the following will be needed:i)incorporating existing use cases(e.g.,those found in 2and 3 and new use cases;and ii)extending the existing network technologies so they can support reliable,low-latency,ubiquitous nature of the involved use cases.conventional sense for storing the vehicle while notin use.Rather,as noted in 8,“garages might becomeservice centres where shared battery-powered cars could becleaned,repaired and recharged before being sent back onthe road.”While this example might seem far into the future,most ofthe technology needed to enable it(high precision maps,realtime traffic information,sensors inside the vehicle such as radars,cameras,ultrasonic,etc.)are either already available or will bein the near future.The most prominent missing component ishigh reliability,low latency communications system.Convergenceof communication technologies with advanced sensors insidethe vehicle,combined with ubiquitous network connectivity andavailable traffic information data,can give rise to seamlessmobility between users origin and destination,without the needto pick up,drop off,or even own a vehicle.The benefits of suchmobility paradigm are multiple:Due to better road utilization and system-level traffic co-ordination,the road capacity is improved,having as aconsequence reduced congestion,better fuel economy,andless time spent in traffic.The user goes directly from his doorstep to the destination,without needing to go to through the intermediate steps of:i)going to pick up the car in a parking;ii)driving to anotherparking instead of the destination;iii)walking from theparking to the destination.The on-demand vehicle also increases the vehicle utilization instead of being idle while the single owner is not using it,the vehicle is being used by multiple users(akin to existingcar-sharing services DriveNow and Car2Go owned by BMWand Daimler,respectively).Since vehicles are utilized more efficiently(i.e.,the idletime is reduced)and combined with automated smart parkinglots 5,the number of parking spots required in the systemcan be decreased considerably.This brings a clear economicand ecological benefit,since parking lots account for thesingle biggest land mass use in large developed cities 4.Some estimates in large cities point out that up to 30%oftraffic on the streets is generated by drivers searching forparking(i.e.,vehicles that already arrived at the destina-tion)4;the new mobility paradigm has the potential tocompletely remove the additional road traffic generated bysearching for parking.In addition to increased efficiency,through the combinationof cooperation and automated driving,the new mobilityparadigm would lead to accident free driving(or,at mini-mum,eradicate the accidents caused by human driver errors).Future vehicle can be perceived as a powerful mobile devicethat is equipped with various sensors(camera,radar,ultra-sound,etc.),having adequate computational resources to support a widerange of automotive services.Many current efforts regardingautomated cars adopt a non-cooperative approach for automateddriving,wherein the key goal is to replace human driver witha“robot”driver:a combination of sensors and software in thevehicle.However,while replacing a human with a robot drivermight have a limited impact in terms of safety,the impact onroad throughput is minimal 1.On the other hand,connectedvehicles(either automated,tele-operated,or human driven)have3the potential to cooperate in order to improve traffic flow onhighways 1,in intersections 9,and in parking lots 5,increasesafety(by seeing around the cornercapability unparalleled byother sensors)and reduce energy consumption 1.In order to achieve the vision described above,a set of enablinguse cases needs to be implemented,many of which rely heavily onthe communication system connecting vehicles with other nearbyvehicles and with infrastructure/backend.In this paper we analyzethese use cases and their requirements and analyze to whatextent the existing technologies can satisfy these requirements.Furthermore,we summarize an architecture for V2X radio accessnetwork,which provides native support for the requirements thatthe various V2X use cases have,most notably low latency,highreliability,and scalability.The rest of the paper is organized as follows.Section IIdescribes relevant V2X use cases and analyzes their requirements.Section III describes communication technologies for enablingV2X.Section IV discusses to which extent a communicationtechnology can support each use case,along with the resulting gapanalysis.It also describes the network architecture and indicateskey technologies for enabling 5G V2X.Section V concludes thepaper.II.USECASES ANDREQUIREMENTSV2X use cases focus on safety,traffic efficiency,and infotain-ment services.Key functional and performance requirements forsafety have already been described by the European Telecommu-nications Standards Institute(ETSI)Intelligent Transport Systems(ITS)2,the US Department of Transportation,and individ-ual research projects.These use cases are used for warningand increasing the environmental awareness based on periodic(e.g.,Cooperative Awareness Message CAM 10)or event-driven(e.g.,Decentralized Environmental Notification Message DENM 11)broadcast messages,with repetition rate as highas 10 Hz(e.g.,emergency vehicle warning)or lower(e.g.,roadworks warning)2.V2X use cases have also been identified by3GPP 12,taking into account services and parameters definedin the first release of ETSI ITS 2.In this group of use casesthe maximum tolerable latency is 100 ms,while the target radiolayer message reception reliability is 95%.These use casesassume a single enabling technology,namely cellular based V2Xcommunication.Enhanced V2X(eV2X)use cases 12 have beendefined by 3GPP as part of Release 15,including more advanceduse cases such as cooperative intersection control,lane merging,and platooning(Fig.2),which have more stringent requirements.The benefits that V2X communication brings to these usecases compared to the sensor-only solutions are multiple.Incase of platooning and Cooperative Adaptive Cruise Control(CACC,Fig.2(a)and lane merging(Fig.2(b),the benefitcomes from proactive communication of locations,speeds,andtrajectories,which results in shorter time and space required toexecute a maneuver.In case of connected automated parkinguse case(Fig.2(c),communication enables centralized controland planning.Furthermore,combined with on-board sensors,ithelps reduce the required lateral and longitudinal spacing betweenvehicles.As noted in 5,this approach can halve the space neededto park a vehicle.Similarly,cooperative intersection control(Fig.2(d)can enable more efficient intersection operation byinforming the vehicles about signal phase and timing,dynamicallyadjusting the speed,and coordinating the flow through multipleintersections based on the current traffic conditions.Due to spacelimitations,we do not provide further details on each of the usecases;rather,we focus on groups of use cases and the overallrequirements of those groups(Table.I).For a detailed description,we refer the reader to 3GPP technical reports 3,12 and a recentoverview article 13.With the increasing availability of vehicles that are capableof supporting higher automation levels,the need for coordi-nation among vehicles and their capability to do so becomesincreasingly more relevant.All automated vehicles rely on thepremise that they continuously plan their trajectories and,basedon the observed environment,select the driving trajectory.Dueto the safety requirements,automated driving sets the moststringent performance requirements for the communication layerin terms of delay,reliability,and capacity.Cooperative LaneChange,Cooperative Collision Avoidance,Convoy Management(Platooning)are typical examples of V2X use cases that areeventually expected to lead to fully connected automated vehicles.The involved vehicles trigger a specific use case for safety reasons(e.g.,emergency maneuver)or for efficient traffic flow(e.g.,platooning)using the monitoring data from the installed sensors,together with the information received from neighboring vehicles.Thereinafter,the automated vehicles undertake to coordinate andplan their maneuvers or trajectories in order to address thetriggering event,based on the built environmental perception.As vehicles advance towards higher automation levels and needto deal with increasingly complex road situations,there will bea need for a complementary communication technology for theexchange of cooperative information with higher bandwidth andimproved reliability.In connected automated vehicles the perfor-mance requirements are more stringent,with certain use casesrequiring ultra-reliable communication links(99%14),withmuch lower maximum end-to-end latency(1-10 ms),and higherdata rate.An initial analysis of Automated Driving use cases hasrecently started with the second release of ETSI ITS specification,while the conclusions of the NGMN 14 and other initiativessummarized in the 5G PPP white paper on automotive sector3have been taken into consideration for performance requirementsderivation.Vehicles will be also connected to one or more ITS appli-c