文章-In-Band-Device-to-Device-Communication-in.docx

A comparison of Bluetooth, WiFi direct and in-band D2D (LTE direct) is provided in Table I. In general, use of an unlicensed spectrum for D2D communication may result in poor quality of service (QoS), because of uncontrolled interference. In addition, the discovery process and setup of connection in out-band D2D or in WiFi direct/Bluetooth is quite complicated since it needs user intervention to establish the connection between the two devices. In contrast to these technologies, inband D2D uses licensed bands, where the above-mentioned problems are solved in a more efficient way. The reason is that the cellular network commonly controls all D2D communication that is currently ongoing and adopts sophisticated allocation and interference mitigation techniques to provide QoS to its users. On the other hand, since cellular networks are managed by the operators, users are likely to have to pay for the connection.
III. D2D CLASSIFICATION
A. D2D Management
From the management point of view, D2D can be classified according to the level of the network’s involvement in control of D2D and a how D2D communication is established (denoted as D2D discovery). In this section, the main characteristic of both management aspects are described and their pros and cons are discussed.
从管理的角度，用户可根据控制D2D网络的参与水平和如何建立D2D通信的分类（记为D2D发现）。在这一部分中，这两个管理方面的主要特点进行了描述和讨论他们的优点和缺点。

1) D2D Control: The D2D control indicates how deeply the network is involved in the control and management of D2D communication. As shown in Fig. 2, the control of D2D can be fully managed by the network (full control), partly managed by the network (loosely control) or hybrid (between full and loosely control).
1） 控制：D2D D2D控制指示如何深入网络参与D2D通信控制和管理。如图2所示，D2D的控制可以通过网络全面管理（完全控制），部分管理由网络（松散控制）或混合（在完全和松散的控制之间）。

In case of full control, the D2D communication is managed by the network of the operator [10]. To be more specific, the network is fully responsible for D2D authentication process during D2D discovery and initiation phase, it handles D2D connection, and allocates power and radio resources. An advantage of the full control approach is that the network can easily coordinate D2D and cellular communications. Thus, the network can mitigate harmful interference between the CUEs and the DUEs. Moreover, the eNB can effectively perform radio resource management and give priorities to individual transmissions to fulfill various QoS requirements.On the other hand, disadvantage of the full control can be seen in high signaling overhead necessary to manage the D2D underlying communication. For example, the eNB has to know the channel state information (CSI) of all involved links for interference avoidance technique and its exchange is very demanding in terms of signaling [32].
在完全控制的情况下，D2D通信是通过运营商的网络管理[ 10 ]。更具体地，D2D的发现和在初始阶段网络全面负责用户认证过程，它处理的D2D连接，并分配电力和无线电资源。该控制方法的优点是，网络可以很容易地协调D2D系统与蜂窝通信。因此，网络可以减轻干扰的线索和会费之间的有害干扰。此外，eNB可以有效执行无线资源管理和优先传输，满足个人不同的QoS要求。另一方面，整个控制的缺点可以在高的信令开销需要管理的底层通信看到D2D。例如，eNB需要知道信道状态信息（CSI）涉及的所有环节的干扰避免技术和交流是非常苛刻的信号[ 32 ]条款。

If loose control is applied, D2D terminals can autonomously communicate between themselves with very limited or no intervention from the network’s side [10]. The network itself is practically responsible only for authentication of the terminals during network entry. The D2D communication can be initiated autonomously and resource allocation or power control is handled solely by the DUEs. The most prominent advantage of this approach consists in generation of only negligible signaling overhead comparing to full control. The critical issue of loosely controlled D2D, however, is the interference caused by the D2D to the legacy CUEs. Consequently, this approach can be avoided by the operators and mobile providers, since the top priority is to guarantee QoS for the CUEs. One possible way to make loosely controlled D2D more feasible is to use unlicensed spectrum for this kind of communication and sharing the frequencies with WiFi or Bluetooth as suggested in [10].
如果采用松动控制，D2D终端可以自主交流彼此之间非常有限的或没有从网络侧[ 10干预]。网络本身实际上只负责在网络进入过程中的终端的身份验证。D2D通信可以启动自主和资源分配和功率控制是由费处理。这种方法的最突出的优点是，在生成的唯一可以忽略不计的信令开销，比较全面的控制。关键的松散控制D2D，然而，由D2D造成的遗留线索的干扰。因此，这种方法可以避免的运营商和移动运营商，因为最优先的是保证QoS的线索。使松散的控制下的D2D更可行的一个可能的方法是使用未授权频谱的这种沟通和WiFi或蓝牙在[ 10建议共享频率]。

The last option for D2D control is a hybrid one [33]. In the hybrid control, the most critical aspects are done by the network (similarod trade-off for the operators, as generated signaling overhead is reasonable, while QoS offered toly as in full control). Among these are authentication process, allocation of radio resources in a large time scale, put restriction on maximal power control allowed at the side of D2D, etc. At the same time, the DUEs are able to manage radio resources and schedule their own transmission and set power control autonomously in a distributive manner according to a short time measurement (similarly as in loosely control). As a consequence, the hybrid control could be seen as a go the CUEs can be guaranteed.
D2D控制最后的选择是一个混合的一个[ 33 ]。在混合控制中，最关键的方面是由网络（类似于在完全控制）。其中，认证过程，在一个大的时间尺度的无线资源分配，将限制在D2D侧允许最大功率控制，同时，会费能够管理无线资源和安排自己的传输和功率控制的自主分配的方式，根据测量时间短（同样作为松散的控制）。作为一个后果，混合控制可以被看作是一个很好的权衡的运营商，所产生的信号开销是合理的，而QoS提供的线索可以保证。

From the above-mentioned, it could be derived that the D2D control influences the amount of signaling needed to be exchanged between the network and the DUEs during ongoing D2D communication. Besides, the D2D control has an impact on the implementation complexity of D2D and it defines how much the UE has to be modified in order to support D2D functionality (in case of loosely controlledD2D, moremodifications are expected at the UE’s side).
综上所述，可以得出，D2D控制影响量的信号所需要的网络和持续的D2D通信在会费之间交换。此外，D2D控制对D2D实现复杂度的影响，它定义了UE进行修改以支持D2D功能多少（在松散controlledd2d，案例moremodifications预计在UE侧）。

2) D2D Discovery: Essential part of the D2D management is a discovery of the DUEs (in literature also known as a peer discovery procedure). The purpose of D2D discovery process is to find the presence of devices that could potentially communicate directly. The overall discovery process can be split into two stages: discovery initiation and discovery control.
2） D2D的发现： D2D管理必不可少的一部分，是一个发现的会费（文献中也称为对等点发现程序）。D2D发现过程的目的是发现设备可能存在的直接沟通。整个发现过程可以分为两个阶段：发现启动和发现控制。

The D2D discovery can be initiated either before the DUEs start to communicate (labeled as “priori”) or during ongoing communication (known as “posteriori”) [34]. The common use of priory D2D discovery is a sharing of a specific content between two devices. On the other hand, posteriori D2D discovery could be used, for example, by mobile devices that move to vicinity of each other during data exchange and D2D communication becomes more suitable/efficient.
D2D发现可以在会费开始交流（标记为“先验的”）或在正在进行的通信（称为“后验”）[ 34 ]。优先级D2D发现常用的两个设备之间共享一个特定的内容。另一方面，后验D2D发现可以使用，例如，通过移动设备，移动到对方附近，在数据交换和D2D通信变得更适合/高效。

Like overall D2D control (described in Section III-A1), the discovery process can be controlled with different levels of network involvement. Thus, the discovery can be controlled either fully by the network (network assisted D2D discovery) or autonomously by the DUEs, which find potential counterparts on its own (autonomous D2D discovery). The network assisted D2D discovery is more convenient as the network is aware of approximate device locations (e.g., whether the DUEs about to communicate with each other are within the same cell or not). On the other hand, network assisted D2D discovery can result in higher signaling overhead due to its centralized nature. The advantage of autonomous D2D discovery is that it has low signaling overhead because it is fully distributed. Nevertheless, discovery process itself could drain battery significantly as it is fully handled by individual devices.
像整体D2D控制（第iii-a1描述），发现过程可以不同程度的网络参与控制。因此，发现可以通过网络完全控制（网络辅助用户发现）或自主的会费，并找到自己的潜在对手（自治D2D发现）。网络辅助的D2D发现更方便的网络意识到近似装置的位置（例如，是否会费要互相沟通是在同一个细胞或不）。另一方面，网络辅助用户发现可以导致更高的开销，由于它的集中性。自主D2D发现的优点是它具有较低的开销，因为它是完全分布式的。然而，发现过程本身可以消耗电池显着，因为它是由个别设备完全处理。

A procedure for network assisted D2D discovery is proposed in [35]. Firstly, the packet data network gateway (P-GW) detects potential D2D candidates. Secondly, a message exchange involving the MME, the eNB, and the UEs participating in D2D is initiated. After the D2D bearer is established, D2D communication can take place instead of conventional cellular communication. The proposed D2D discovery, however, can result in quite significant signaling overhead. Another network assisted D2D discovery utilizing resources allocated for the device discovery is proposed in [36]. At a specific discovery interval, several UEs perform discovery by means of a discovery message. To avoid contention among them, the UEs take three steps, which are random selection of; i) search/listen state, ii) discoverable interval, and iii) frequency multiplexed discovery channel. The results indicate that the proposed technique is able to increase the amount of discovered D2D within one discovery period. Signature-based D2D discovery, during which the DUEs transmit discovery signal using temporary identity, is proposed in [37]. The paper shows how discovery signal is mapped to the physical resources and how to avoid collisions at discovery channel, which is allocated by the network.
一种网络辅助D2D发现程序在[ 35 ]提出了。首先，分组数据网网关（P-GW）检测到潜在的D2D的候选人。其次，包括MME，eNB的消息交换，并参与D2D UE发起。在D2D承载的建立，D2D通信可以取代传统的蜂窝通信。提出了D2D的发现，但是，可能会导致非常重要的信令开销。另一个网络辅助D2D发现利用分配给设备发现资源是[ 36种]。在一个特定的发现间隔，几个问题执行发现的发现意味着。为避免冲突，其中的问题，采取三个步骤，即随机选择；我）搜索/听状态，二）发现的间隔，和III）发现信道频率复用。结果表明，该技术能够增加发现D2D在一发现时期。基于D2D发现签名，其间发现信号传输使用会费临时身份，在[ 37 ]提出了。本文展示了如何发现信号被映射到物理资源，以及如何避免在发现信道，这是由网络分配的碰撞。

Autonomous D2D discovery technique intended for communication systems based on Qualcomm’s defined D2D—FlashLinQ [38] is introduced in [39]. A fully distributed D2D discovery for synchronous OFDM-based system using time division multiplexing (TDM) technique is proposed in [40]. The radio resources contain discovery region, during which the devices receive or transmit discovery signals. This allows each device to advertise its presence and service and to discover other close devices autonomously and continuously in distributed manner. Another autonomous D2D discovery protocol based on dynamic source routing protocol is introduced in [41].The proposed scheme floods the network with modified discovery packets. These packets include channel number, power used for transmission and interference power measured by transmitting mode. Hence, the receiver is able to calculate the path loss and signal to interference plus noise ratio (SINR), which are used for estimation of minimum transmit power to be heard by the transmitter. As a result, two nodes can create D2D pair if bidirectional link can be established while the constraint on power is set to minimize interference to the CUEs. Using peer discovery resources for beacon signal is proposed in [42]. To ensure low discovery overhead, only small part of physical layer frames are used for the D2D discovery. The discovery of neighboring device is determined according to SINR measured from the received beacon.
自治D2D发现技术用于基于高通的定义的D2D通信系统flashlinq [ 38 ]了[ 39 ]。一个完全分布式的D2D发现使用时分复用（TDM）的OFDM系统的同步技术是在[ 40 ]提出了。无线资源包含发现区域，在此期间，该设备接收或发送发现信号。这允许每个设备来宣传它的存在和服务，并发现其他封闭的设备，自主地和不断地在分布式的方式。另一个自治D2D发现协议基于动态源路由协议进行了[ 41 ]。该方案的洪水与改性发现分组网络。这些数据包包括信道数，用于传输和通过发射模式测量的干扰功率的功率。因此，接收机能够计算路径损耗和信号干扰噪声比（SINR），这是用于估计的最小发射功率是由发射机听到。因此，如果两个节点的双向链路可以建立在权力的约束将减少干扰的线索建立D2D对。在[ 42 ]中提出了利用对等点发现的信标信号的资源。为了保证低发现开销，只有小部分的物理层帧用于D2D的发现。周边设备的发现是根据信噪比从接收到的信标测定。

B. D2D Communication Scenarios
This section illustrates individual possible scenarios that could be considered for D2D communication. In general, D2D scenarios can be classified by several aspects (see Fig. 2) as follows.
本节说明了个体在D2D通信中的可能的情况。总的来说，D2D场景可以由几个方面分类（见图2）如下。

• Coverage—This aspect distinguishes whether the pair of DUEs is under the coverage of a cellular network. In this context we can categorize D2D communication as:
覆盖这一方面区分了这一对会费是否在蜂窝网络的覆盖范围内。在这种情况下，我们可以将D2D通信：
— In coverage—Both DUEs are within the coverage of the cellular network.
这两个费用都在蜂窝网络的覆盖范围内。
— Partial coverage—One DUE is in the coverage of the cellular network whereas the second one is out of coverage (e.g., it could be in a coverage hole caused by interferes in the proximity).
一个是由于在蜂窝网络的覆盖范围，而第二个是出的覆盖范围（例如，它可能是在一个覆盖孔引起的干扰在接近）。
— Out of coverage—Both DUEs are outside the cellular communication network. Note that this scenario is considered mainly in 3GPP for public safety cases,when the network can be temporarily disabled.
这两种费用都在蜂窝通信网络的外部。注意，这种情况主要是在公共安全案件被3GPP，当网络可以暂时禁用。

• Type of D2D communication—This aspect expresses how many DUEs are involved in D2D communication:
这一方面表达了多少会费参与D2D通信：
— One-to-one communication—Direct communication between two DUEs that creates one D2D communication pair.
— One-to-many communication—One DUE multicasts/broadcasts data to several DUEs in a cluster.This option is also labeled device to multi-device (D2MD).
由于组播/广播集群中的几个费数据。此选项也标有多装置（d2md）。

• Area of D2D communication—The aspect distinguishing whether both communicating DUEs are served by the same cell or not:
区分是否两个通信的会费由相同的单元格或不区分：
— The same cell—The DUEs creating a D2D pair or a cluster are attached to the same eNB.
会费创造了D2D对或一簇附着在同一个基站。
— The different cell—The DUEs belonging to the same D2D pair or cluster are attached to different eNBs.
会费属于相同的D2D对或集群连接到不同的基站。

• Relaying functionality—The DUE may have relaying functionality to retransmit data of other DUEs within its proximity. This feature can be used to:
由于可能有中继功能发送其他数据在其邻近的会费。此功能可用于：
— Enhance capacity - The DUE attached to another DUE with relay functionality is usually in coverage of the eNB.
由于连接到另一个由于继电器的功能通常是在eNB覆盖。
— Extend coverage - The DUE out of coverage may use other DUEs in order to reach the eNB.
由于从覆盖可以使用为其它费用达到ENB。

Depending on the above-mentioned four aspects, severalscenarios for D2D communication can be specified (see Fig. 3). First, a set of D2D scenarios corresponds to those taken into account in a 3GPP standardization group, according to [22]. Note that scenario labeling is in line with 3GPP and also we consider only scenarios when all eNBs belong to the same public land mobile network (PLMN). In addition to 3GPP scenarios, we include possible options for D2D communication tackled in research studies but not currently standardized in 3GPP.
根据上述四个方面，对D2D通信的几种场景可以指定（见图3）。首先，一组D2D方案对应那些在3GPP标准化组考虑，根据[ 22 ]。注意场景标签是符合3GPP和我们也只考虑方案时都已经属于同一公共陆地移动网（PLMN）。除了3GPP方案，我们包括D2D通信解决的研究但在3GPP目前没有标准化的可能的选择。

In general, 3GPP categorizes D2D scenarios into two groups;without relays and with relays. The scenarios without relaying are more conventional, since there are no special requirements imposed on the UE in order to support relaying functionality.The most common scenario used in contemporary studies and standards is a simple communication between two devices underlying cellular communication, where both devices are served by the s