08中英文雙語(yǔ)畢業(yè)設(shè)計(jì)外文文獻(xiàn)翻譯成品超密集設(shè)備對(duì)設(shè)備通信的資源分配方案--5g蜂窩網(wǎng)絡(luò)

1、<p>　　A Resource Allocation Scheme for Device-to-Device Communication over Ultra-Dense</p><p>　　5G Cellular Networks</p><p>　　外文標(biāo)題：A Resource Allocation Scheme for Device-to-Device Communic

2、ation over Ultra-Dense 5G Cellular Networks</p><p>　　外文作者：J Gu ， HW Yoon ， J Lee ， SJ Bae ， YC Min</p><p>　　文獻(xiàn)出處：《Applied Mechanics & Materials》 , 2015 , 713-715 :1208-1215</p><p&

3、gt;　　英文2655單詞，15098字符，中文3509漢字。</p><p>　　此文檔是外文翻譯成品，無(wú)需調(diào)整復(fù)雜的格式哦！下載之后直接可用，方便快捷！只需二十多元。</p><p><b>　　Abstract</b></p><p>　　With the advance of the wireless technologies and

4、mobile devices, the number of mobile devices in cellular system will explosively grow. According to this trend, a base station will face a heavy traffic loading and even cannot provide services for a large amount of mobi

5、le devices. To deal with this issue, device-to-device (D2D) technology is a promising solution to increase spectrum efficiency by reusing radio resource blocks. In this paper, we study a resource allocation problem with

6、the object</p><p>　　Index Terms—D2D, cellular networks, resource allocation</p><p>　　I. INTRODUCTION</p><p>　　With the advance of mobile devices, mobile data traffic will explosivel

7、y grow every year. In other words, each base station requires to serve more users and suffers a problem that radio resources are not sufficient to provide services for ultra-dense users. Device-to- device (D2D) communica

8、tions is one of the solutions to offload the traffic loading of a base station. A pair of D2D devices can reuse a frequency, used by another pair of D2D devices, to increase spectrum efficiency and system capa</p>

9、<p>　　Recently, several researches have studied radio resource allocation problem for D2D communications and can be classified into two directions. The first one considers that the number of cellular users (CUEs) i

10、s more than the number of D2D users (DUEs). [1] proposed a resource allocation algorithm that relies on an area controller and can be used for uplink and downlink. [2–6] further considered mode selection that a base stat

11、ion can select cellular transmission mode or D2D mode for a user. After </p><p>　　papers propose resource allocation algorithms with different considerations to allocate radio resource for each user.</p&g

12、t;<p>　　However, the above works do not consider a scenario that the</p><p>　　number of cellular users is less than the number of D2D users so that the radio resources cannot be efficiently reused for

13、 a large amount of D2D users. Specifically, above works restrict that radio resource blocks used by a pair of D2D user can only be reused by a pair of D2D users.</p><p>　　Although [2, 3] do not have the rest

14、riction, the time complexity of the proposed algorithms is extremely high and the proposed algorithm cannot be used in an ultra-dense environment.</p><p>　　The second direction considers that the number of c

15、ellular users is less than the number of D2D users. [7] considered that users in a region cannot use the same radio resource block to avoid interference and proposed a coloring method to allocate radio blocks for each us

16、er. The related works firstly allocate radio resource blocks for cellular users and then considers D2D users to reuse the radio resource blocks. However, we observes that if we firstly allocate radio resource blocks for

17、cellular </p><p>　　In this paper, we study a radio resource block allocation problem in an ultra-dense 5G cellular system and consider that the number of D2D users is more than the number of cellular users.

18、The objective is to maximize the system capacity. Then, we propose a coloring algorithm to allocate radio resource blocks to each D2D and cellular user. The proposed algorithm will firstly allocate radio resource blocks

19、to D2D users and then to cellular users such that each radio resource block can be efficient</p><p>　　II describes system model. In Section III, we develop algorithms to solve the problem. Simulation results

20、 are presented in Section IV. Finally, Section V conclusions this paper.</p><p>　　II. SYSTEM MODEL</p><p>　　In this paper, we consider that there is a base station to serve a set C of cellular

21、users (CUEs) and a set D of D2D users (DUEs). DT is used to represent the set of DUEs which transmits data and DUE d ?DT means that the DUE requires to transmit data. DR is used to represent the set of DUEs which receive

22、s data and d?DR means that the DUE has to receive data. The base station has to determine that each user should operate on cellular mode or D2D mode. Then, the base station should allocate radio</p><p>　　whe

23、re PC is the transmit power of a CUE and PD is the transmit power of a DUE. GB,c is the channel gain between the base station and CUE c and Gd,d’ is the channel gain between DUE d and DUE d ’. N0 is the additive white Ga

24、ussian noise. The total data rate for all CUEs can be expressed as</p><p>　　where W is the bandwidth of a RB. </p><p>　　When RB r is allocated to DUE d, the SINR for DUE d can be expressed as<

25、;/p><p>　　The total data rate for all DUEs is</p><p>　　Objective: Our objective is to determine that each user should operate on cellular or D2D mode and decide that each RB should be allocated to

26、which users such that the total system capacity can be maximized. The objective function is expressed as</p><p>　　III. ALGORITHM DESCRIPTION</p><p>　　In a metropolitan environment with high dens

27、ity characteristics, our scheme is different from the traditional communication model and uses a large number of D2D communication mode to replace the original cellular communication mode.Thus, our scheme not only effect

28、ively reduces the load on the base station (BS). The device also reduces power consumption and increases data transmission rate. In addition, our scheme is no restrictions on the sharing of resources and chooses the best

29、 resource alloca</p><p>　　To allocate the resource, if the CUE first selects the RB, it affectsthe surrounding DUEs, reduces the choice of resources, and even causes no resources to be chosen by DUEs. Theref

30、ore, the DUE has higher priority to obtain the RB so that it reduces the CUE selection resources to increase the spectrum utilization. Our proposed scheme is divided into two stages. In the first stage, it is the mode al

31、location algorithm. In the second stage, it is the resource selection algorithm. The proposed schem</p><p>　　If neighbor devices have the needed information for a new device, a new device directly uses the D

32、2D mode to obtain the information from the neighbor device. Contrarily, if neighbor devices do not have the needed information, the system determines the mode according to the resource usage. In other words, if the resou

33、rce of the system does not be used by otherCUEs, the new device uses cellular mode.Otherwise, the system rejects the request of the new device.Our scheme allows the device to use the</p><p>　　Resource select

34、ion algorithm is also divided into two stages. In thefirst stage, we establish ”neighbor graph”. Neighbor graph is defined as the relationship among devices. In the neighbor graph, a node represents a device, and an edge

35、 connects two nodes which use the same resource block (RB) to cause excessive interference each other. The color represents the RB, and two nodes with the edge connected do not use the same color. The second stage is the

36、 selection of resources. It uses previously e</p><p>　　In order to find the best l for establishing the neighbor graph, weuse the limitation of the graph coloring as the upper bound of the radius l and calcu

37、late the minimum acceptance value of SINR as lower bound. Hence, we use l to determine whether devices are the neighbor or not. When the distance between two devices are less than l, we set two devices as neighbors. All

38、nodes establish their neighbor graph in turn and calculate the number of neighbors. When the number of neighbors exceeds the upp</p><p>　　Next, based on the graph coloring, the BS can use the established nei

39、ghbor graph, and adjacent devices cannot use the same color to select the RB. When we allocate the RB, the DUE first selects resources and then CUE selects resources. Initially, each device removes RBs which are used by

40、neighbor devices from the resource candidate list, and the CUE deletes RBs that have been used by other CUEs. </p><p>　　After sorting, each device calculates the SINR for each candidate RB</p><p&g

41、t;　　in order to select the suitable RB. In the sorting process, when two RBs provide the same SINR, we consider the current situation of the RB. In addition to achieving the highest system performance, we also improve th

42、e spectrum utilization. Thus, we use the number of users to select the RB, which is used by more devices. Users using the same RB can maintain above the acceptable SINR value as much as possible so that more users choose

43、 the same piece of RB to enhance the spectrum usage. Algorithm 3</p><p>　　IV. SIMULATION SETUP</p><p>　　We developed a simulation model to evaluate the performance of our proposed algorithm, den

44、oted as mode allocation and resource selection (MARS) algorithm. The proposed algorithm is compared with an algorithm proposed in [7]. The baseline is denoted as graph- coloring resource allocation (GOAL). The simulation

45、 settings are listed in Table I.</p><p>　　SIMULATION RESULTS</p><p>　　Fig. 1 shows the impacts of the number of D2D pairs on the total system capacity. When the number of D2D pairs increases, th

46、e total system capacity increases. The result is expected because more users can contribute more data rate when RBs can be efficiently reused by D2D users. Our proposed algorithm can increase more system capacity than th

47、e baseline when there are more D2D users. This is because our proposed algorithm considers ultra-dense environment and firstly allocates RBs for D2D users s</p><p>　　VI CONCLUSION</p><p>　　This

48、paper studies the resource allocation problem over ultra-dense5G cellular systems and considers that the number of D2D users is more than the number of cellular users. The objective is to maximizethe system capacity. Thi

49、s paper observes that RBs should be firstly allocated to D2D users when the number of D2D users is more thanthat of cellular users such that the RBs can be more efficiently reused. To solve the target problem, this paper

50、 proposes a mode selection todetermine that each user s</p><p>　　REFERENCES</p><p>　　[1] F. Malandrino, Z. Limani, C. Casetti, and C. F. Chiasserini,</p><p>　　“Interference-Aware Do

51、wnlink and Uplink Resource Allocation in</p><p>　　HetNets With D2D Support,” IEEE Transactions on Wireless</p><p>　　Communications, vol. 14, no. 5, pp. 2729–2741, May 2015.</p><p>　

52、　[2] L. Lei, Y. Kuang, N. Cheng, X. S. Shen, Z. Zhong, and C. Lin,</p><p>　　“Delay-Optimal Dynamic Mode Selection and Resource Allocation in</p><p>　　Device-to-Device CommunicationsPart I: Optim

53、al Policy,” IEEE</p><p>　　Transactions on Vehicular Technology, vol. 65, no. 5, pp. 3474–3490,</p><p><b>　　May 2016.</b></p><p>　　[3] ——, “Delay-Optimal Dynamic Mode Sel

54、ection and Resource</p><p>　　Allocation in Device-to-Device CommunicationsPart II: Practical</p><p>　　Algorithm,” IEEE Transactions on Vehicular Technology, vol. 65, no.</p><p>　　5,

55、 pp. 3491–3505, May 2016.</p><p>　　[4] K. Zhu and E. Hossain, “Joint Mode Selection and Spectrum</p><p>　　Partitioning for Device-to-Device Communication: A Dynamic</p><p>　　Stackel

56、berg Game,” IEEE Transactions on Wireless Communications,</p><p>　　vol. 14, no. 3, pp. 1406–1420, March 2015.</p><p>　　[5] J. Zheng, R. Chen, and Y. Zhang, “Dynamic Resource Allocation</p>

57、<p>　　based on Service Time Prediction for Device-to-Device</p><p>　　Communication Underlaying Cellular Networks,” IET</p><p>　　Communications, vol. 9, no. 3, pp. 350–358, Feb. 2015.</

58、p><p>　　[6] H. Zhang, L. Song, and Z. Han, “Radio Resource Allocation for</p><p>　　Device-to-Device Underlay Communication Using Hypergraph</p><p>　　Theory,” IEEE Transactions on Wirel

59、ess Communications, vol. 15,</p><p>　　no. 7, pp. 4852–4861, March 2016.</p><p>　　[7] X. Cai, J. Zheng, and Y. Zhang, “A Graph-Coloring based</p><p>　　Resource Allocation Algorithm f

60、or D2D Communication in Cellular</p><p>　　Networks,” IEEE International Conference on Communications</p><p>　　(ICC), pp. 5429–5434, September 2015.</p><p>　　超密集設(shè)備對(duì)設(shè)備通信的資源分配方案--5G蜂窩網(wǎng)

61、絡(luò)</p><p><b>　　摘要</b></p><p>　　隨著無(wú)線技術(shù)和移動(dòng)設(shè)備的進(jìn)步，蜂窩系統(tǒng)中的移動(dòng)設(shè)備數(shù)量將爆炸式增長(zhǎng)。根據(jù)這一趨勢(shì)，基站將面臨沉重的流量負(fù)載，甚至無(wú)法為大量移動(dòng)設(shè)備提供服務(wù)。為了解決這個(gè)問(wèn)題，設(shè)備到設(shè)備（D2D）技術(shù)是通過(guò)重新使用無(wú)線資源塊來(lái)提高頻譜效率的有希望的解決方案。在本文中，我們研究了一個(gè)資源分配問(wèn)題，目標(biāo)是在超密集5G蜂窩系統(tǒng)

62、上最大化系統(tǒng)容量，并考慮D2D用戶數(shù)量多于蜂窩用戶的情況。本文觀察到無(wú)線資源塊應(yīng)該首先分配給超密集場(chǎng)景下的D2D用戶。然后，我們提出資源分配方法來(lái)解決這個(gè)問(wèn)題。仿真結(jié)果符合我們的觀察，并表明該方案可以顯著提高系統(tǒng)容量和頻譜效率</p><p>　　索引術(shù)語(yǔ) - D2D，蜂窩網(wǎng)絡(luò)，資源分配</p><p><b>　　一，導(dǎo)言</b></p><p&

63、gt;　　隨著移動(dòng)設(shè)備的進(jìn)步，移動(dòng)數(shù)據(jù)流量將每年爆炸式增長(zhǎng)。換句話說(shuō)，每個(gè)基站需要為更多的用戶提供服務(wù)，并且存在無(wú)線資源不足以為超密用戶提供服務(wù)的問(wèn)題。設(shè)備到設(shè)備（D2D）通信是卸載基站業(yè)務(wù)負(fù)載的解決方案之一。一對(duì)D2D設(shè)備可以重用另一對(duì)D2D設(shè)備使用的頻率，以在兩對(duì)D2D設(shè)備互不干擾時(shí)增加頻譜效率和系統(tǒng)容量。在一個(gè)超密集的環(huán)境中，如何分配無(wú)線資源塊是一個(gè)重要的問(wèn)題，并且顯著影響頻譜效率和系統(tǒng)容量。本文研究了超密集5G蜂窩系統(tǒng)中D2D通

64、信下的無(wú)線資源塊分配問(wèn)題。</p><p>　　最近，一些研究已經(jīng)研究了用于D2D通信的無(wú)線電資源分配問(wèn)題，并且可以分為兩個(gè)方向。第一個(gè)認(rèn)為蜂窩用戶（CUE）的數(shù)量多于D2D用戶（DUE）的數(shù)量。 [1]提出了一種依賴于區(qū)域控制器的資源分配算法，可用于上行鏈路和下行鏈路。 [2-6]進(jìn)一步考慮了基站可以為用戶選擇蜂窩傳輸模式或D2D模式的模式選擇。確定每個(gè)用戶的模式后，</p><p>　

65、　論文提出了具有不同考慮因素的資源分配算法，為每個(gè)用戶分配無(wú)線資源。</p><p>　　不過(guò)，上述作品并不考慮這種情況</p><p>　　蜂窩用戶的數(shù)量少于D2D用戶的數(shù)量，無(wú)法為大量的D2D用戶高效地重用無(wú)線資源。具體來(lái)說(shuō)，上述工作限制了一對(duì)D2D用戶使用的無(wú)線資源塊只能由一對(duì)D2D用戶重新使用。</p><p>　　盡管[2,3]沒(méi)有限制，但所提出的算法的時(shí)

66、間復(fù)雜度非常高，并且所提出的算法不能用于超密集環(huán)境中。</p><p>　　第二個(gè)方向認(rèn)為蜂窩用戶數(shù)量少于D2D用戶數(shù)量。 [7]認(rèn)為，一個(gè)地區(qū)的用戶不能使用相同的無(wú)線資源塊來(lái)避免干擾，并提出了一種著色方法來(lái)為每個(gè)用戶分配無(wú)線塊。相關(guān)工作首先為蜂窩用戶分配無(wú)線資源塊，然后考慮D2D用戶重用無(wú)線資源塊。然而，我們觀察到，如果我們首先為蜂窩用戶分配無(wú)線資源塊，則大量的D2D用戶不能在超密集環(huán)境中有效地重復(fù)使用相同的無(wú)

67、線資源塊。這是因?yàn)榉涓C用戶的發(fā)射功率一般大于D2D用戶的發(fā)射功率，使得大量的D2D用戶會(huì)受到干擾，無(wú)法重用相同的無(wú)線資源塊。</p><p>　　在本文中，我們研究了超密集5G蜂窩系統(tǒng)中的無(wú)線資源塊分配問(wèn)題，并認(rèn)為D2D用戶的數(shù)量超過(guò)了蜂窩用戶的數(shù)量。 目標(biāo)是最大化系統(tǒng)容量。 然后，我們提出了一種著色算法來(lái)為每個(gè)D2D和蜂窩用戶分配無(wú)線資源塊。 該算法首先將無(wú)線資源塊分配給D2D用戶，然后分配給蜂窩用戶，使得每個(gè)

68、無(wú)線資源塊可以被有效地重用。 仿真結(jié)果表明，該方案顯著提高了系統(tǒng)吞吐量和頻譜效率。</p><p>　　本文的其余部分安排如下。 部分</p><p>　　II描述了系統(tǒng)模型。 在第三節(jié)中，我們開發(fā)算法來(lái)解決這個(gè)問(wèn)題。 第四節(jié)給出了仿真結(jié)果。 最后，本文第五節(jié)結(jié)論。</p><p><b>　　I。 系統(tǒng)模型</b></p>&l

69、t;p>　　在本文中，我們認(rèn)為有一個(gè)基站服務(wù)C組蜂窩用戶（CUE）和D組D2D用戶（DUE）。 DT用于表示傳輸數(shù)據(jù)的DUE的集合，DUE表示DUE表示DUE要求傳輸數(shù)據(jù)。 DR用于表示接收數(shù)據(jù)的DUE集合，d?DR表示DUE必須接收數(shù)據(jù)。 基站必須確定每個(gè)用戶應(yīng)該在蜂窩模式或D2D模式下工作。 然后，基站應(yīng)該為每個(gè)CUE和每對(duì)DUE分配無(wú)線資源塊（RB）。 分配給CUE和DUE的資源塊集合分別表示為RC和RD。 當(dāng)rc??RC時(shí)

70、，表示RB r被分配給CUE c。</p><p>　　rd??RD表示RB r分配給DUE d。 rc = rd表示CUE c和DUE d同時(shí)使用RB r。 當(dāng)RB r被分配給CUE c時(shí)，CUE c的信號(hào)干擾比（SINR）可以表示為</p><p>　　PC是CUE的發(fā)射功率，PD是DUE的發(fā)射功率。 GB，c是基站和CUE c和Gd之間的信道增益，d'是DUE d和DUE d

71、'之間的信道增益。 N0是加性高斯白噪聲。 所有CUE的總數(shù)據(jù)速率可以表示為</p><p>　　其中W是RB的帶寬。 當(dāng)RB r被分配給DUE d時(shí)，DUE d的SINR可以表示為</p><p>　　所有DUE的總數(shù)據(jù)速率為</p><p>　　目標(biāo)：我們的目標(biāo)是確定每個(gè)用戶應(yīng)該在蜂窩或D2D模式下運(yùn)行，并決定每個(gè)RB應(yīng)該分配給哪些用戶，以使系統(tǒng)總?cè)萘孔?/p>

72、大化。 目標(biāo)函數(shù)表示為</p><p><b>　　III。算法描述</b></p><p>　　在高密度特性的城市環(huán)境中，我們的方案不同于傳統(tǒng)的通信模式，并且采用大量的D2D通信模式來(lái)取代原有的蜂窩通信模式。</p><p>　　因此，我們的方案不僅有效地減少了基地的負(fù)荷</p><p>　　站（BS）。該器件還可以降

73、低功耗并提高數(shù)據(jù)傳輸速率。此外，我們的方案不受資源共享的限制，并根據(jù)具體情況選擇最佳的資源分配，以增加資源使用的靈活性。因此，整體頻譜利用率也可以得到有效改善。</p><p>　　要分配資源，如果CUE首先選擇RB，它會(huì)影響</p><p>　　周圍的DUE會(huì)減少資源的選擇，甚至不會(huì)導(dǎo)致DUE選擇任何資源。因此，DUE獲得RB的優(yōu)先級(jí)更高，從而減少CUE選擇資源以提高頻譜利用率。我們提出

74、的方案分為兩個(gè)階段。在第一階段，它是模式分配算法。在第二階段，它是資源選擇算法。所提出的方案僅適用于新設(shè)備，以便我們有效減少處理用戶的數(shù)量以縮短計(jì)算時(shí)間。</p><p>　　算法1是模式分配算法。當(dāng)一個(gè)新的設(shè)備</p><p>　　進(jìn)入BS的服務(wù)區(qū)域以發(fā)送服務(wù)請(qǐng)求，BS根據(jù)當(dāng)時(shí)的資源使用情況并參考新設(shè)備的周圍設(shè)備。因此，我們決定使用哪種通信模式用于新設(shè)備。</p><

75、p>　　如果鄰居設(shè)備具有新設(shè)備所需的信息，則新設(shè)備直接使用D2D模式獲取鄰居設(shè)備的信息。相反，如果鄰居設(shè)備沒(méi)有所需的信息，則系統(tǒng)根據(jù)資源使用情況確定模式。換言之，如果系統(tǒng)的資源不被其他CUE使用，則新設(shè)備使用蜂窩模式。否則，系統(tǒng)拒絕新設(shè)備的請(qǐng)求。我們的方案允許設(shè)備使用D2D模式接收數(shù)據(jù)，縮短傳輸時(shí)間，提高整體系統(tǒng)性能。</p><p>　　資源選擇算法也分為兩個(gè)階段。在第一階段，我們建立“鄰居圖”。鄰居圖被

76、定義</p><p>　　作為設(shè)備之間的關(guān)系。在鄰居圖中，節(jié)點(diǎn)表示設(shè)備，邊連接使用相同資源塊（RB）的兩個(gè)節(jié)點(diǎn)彼此造成過(guò)度干擾。顏色表示RB，并且連接邊的兩個(gè)節(jié)點(diǎn)不使用相同的顏色。第二階段是資源的選擇。它使用先前建立的鄰居圖進(jìn)行圖著色來(lái)完成所有設(shè)備的資源選擇。為了建立鄰居圖，如何選擇鄰居圖中節(jié)點(diǎn)的半徑是一個(gè)重要問(wèn)題。如果半徑過(guò)大導(dǎo)致覆蓋區(qū)域內(nèi)有大量設(shè)備，系統(tǒng)無(wú)法完成所有設(shè)備的資源選擇;相反，如果半徑太小而不能導(dǎo)致

77、使用相同的RB來(lái)引起過(guò)多干擾的設(shè)備。因此，本文考慮圖著色和SINR限制的兩個(gè)基本角度，計(jì)算鄰近圖的最佳半徑。</p><p>　　首先，根據(jù)圖形著色，顏色的數(shù)量必須是</p><p>　　大于節(jié)點(diǎn)的數(shù)量。換句話說(shuō)，RB的數(shù)量大于或等于設(shè)備的數(shù)量。因此，我們限制鄰居圖中半徑l的上限，并且只有內(nèi)部設(shè)備的數(shù)量小于RB的總數(shù)。</p><p>　　為了找到建立鄰居圖的最佳l

78、，我們</p><p>　　使用圖著色的限制作為半徑l的上界，并計(jì)算SINR的最小接受值作為下限。因此，我們使用l來(lái)確定設(shè)備是否是鄰居。當(dāng)兩臺(tái)設(shè)備之間的距離小于1時(shí)，我們將兩臺(tái)設(shè)備設(shè)置為鄰居。所有節(jié)點(diǎn)依次建立其鄰居圖并計(jì)算鄰居數(shù)。當(dāng)鄰居數(shù)量超過(guò)l的上限時(shí)，必須重新建立鄰居圖。算法2中顯示了鄰居圖的算法。</p><p>　　接下來(lái)，基于圖著色，BS可以使用建立的鄰居圖，并且相鄰設(shè)備不能使用

79、相同的顏色來(lái)選擇RB。當(dāng)我們分配RB時(shí)，DUE首先選擇資源，然后CUE選擇資源。最初，每個(gè)設(shè)備從資源候選列表中去除由相鄰設(shè)備使用的RB，并且CUE刪除已被其他CUE使用的RB。</p><p>　　分類后，每個(gè)設(shè)備計(jì)算每個(gè)候選RB的SINR以選擇合適的RB。在排序過(guò)程中，當(dāng)兩個(gè)RB提供相同的SINR時(shí)，我們考慮RB的當(dāng)前狀況。除了實(shí)現(xiàn)最高的系統(tǒng)性能外，我們還提高了頻譜利用率。因此，我們使用用戶數(shù)量來(lái)選擇更多設(shè)備使

80、用的RB。使用相同RB的用戶可以盡可能保持在可接受的SINR值以上，以便更多的用戶選擇同一塊RB來(lái)增強(qiáng)頻譜使用。算法3顯示了資源選擇算法的細(xì)節(jié)。</p><p><b>　　IV。 仿真設(shè)置</b></p><p>　　我們開發(fā)了一個(gè)仿真模型來(lái)評(píng)估我們提出的算法的性能，表示為模式分配和資源選擇（MARS）算法。 該算法與文獻(xiàn)[7]中提出的算法進(jìn)行了比較。 基線表示為圖

81、形著色資源分配（GOAL）。 表I列出了仿真設(shè)置。</p><p><b>　　五，仿真結(jié)果</b></p><p>　　圖1顯示了D2D對(duì)的數(shù)量對(duì)總系統(tǒng)容量的影響。 當(dāng)D2D對(duì)的數(shù)量增加時(shí)，總系統(tǒng)容量增加。 預(yù)期結(jié)果是因?yàn)楫?dāng)D2D用戶可以有效地重用RB時(shí)，更多的用戶可以貢獻(xiàn)更多的數(shù)據(jù)速率。 當(dāng)存在更多的D2D用戶時(shí)，我們提出的算法可以比基線增加更多的系統(tǒng)容量。 這是

82、因?yàn)槲覀兲岢龅乃惴紤]超密集環(huán)境，并首先為D2D用戶分配RB，這樣可以更有效地重用RB。</p><p><b>　　六，結(jié)論</b></p><p>　　本文研究了超密集5G蜂窩系統(tǒng)的資源分配問(wèn)題，認(rèn)為D2D用戶的數(shù)量為超過(guò)蜂窩用戶的數(shù)量。 目標(biāo)是最大化系統(tǒng)容量。 本文指出RB應(yīng)該是第一個(gè)當(dāng)D2D用戶的數(shù)量多于蜂窩用戶的數(shù)量時(shí)，分配給D2D用戶，使得RB可以被更有效