LTE IN BULLET学习材料
2.1 INTRODUCTION * Long Term Evolution (LTE) startsfrom 3GPP release 8 * 3GPP Technical Report 25.913defines the key objectives of LTE as: ¡ support for a flexible transmissionbandwidth up to 20 MHz ¡ peak downlink data rate of 100 Mbps whenusing 2 receive antenna at the UE ¡ peak uplink data rate of 50 Mbps when using1transmit antenna at the UE ¡ round trip time of less than 10 ms ¡ downlink average spectrum efficiencyimproved 3 to 4 times relative to release 6 HSDPA ¡ uplink average spectrum efficiency improved2 to 3 times relative to release 6 HSDPA * LTE has a flat architecture whichminimises the number of network elements * LTE is optimised for PacketSwitched (PS) services but includes functionality to handle Circuit Switched(CS) services, e.g. CS fallback to UMTS * LTE supports the speech serviceusing Voice over IP. Otherwise, the speech service can be supported by allowingthe UE to fallback to UMTS, GSM or CDMA2000 * LTE supports Multimedia BroadcastMulticast Services (MBMS) for the transmission of mobile TV * Frequency Division Duplex (FDD)and Time Division Duplex (TDD) versions of LTE have been standardised. Bothallow channel bandwidths of up to 20 MHz * LTE allows inter-working withexisting GSM, UMTS and CDMA2000 technologies * LTE uses QPSK, I6QAM and 64QAMmodulation schemes with OFDMA (downlink) and SC-FDMA (uplink) multiple accesstechnologies * LTE supports Multiple InputMultiple Output (MIMO) antenna technology in the downlink direction. 3GPPreleases 8 and 9 do not support MIMO in the uplink direction * Existing spcctrum allocations canbe re-farmed for the introduction of LTE * LTE simplifies network planning byminimising the requirement for manually planned neighbour lists * LTE includes Self OrganisingNetwork (SON) functionality to help automate network configuration,optimisation, fault finding and fault handling * LTE Advanced starts from 3GPPrelease 10 * LTE Advanced introduces Carrier Aggregationto provide wider effective channel bandwidths. It also introduces MIMO in theuplink direction, as well as increasing the number of antenna elements whichcan be used for MIMO in the downlink direction * Other technologies continue todevelop in parallel to LTE, e.g. UMTS introduces HSPA+ with MIMO, 64QAM andMulti-Carrier Transmission allowing effective channel bandwidths of 10, 20 and40 MHz 2.2 ARCHITECTURE * LTE refers to the Evolved UMTSRadio Access Network (E_UTRAN), Whereas System Architecture Evolution (SAE)refers to the Evolved Packet Core(EPC), Figure 1 illustrates this divisionbetween radio access and core networks * LTE uses a flat architecturewithout a Radio Network Controller (RNC),nor Basc Station Controller (BSC) * The LTE equivalent of a UMTS NodeB is an ‘evolved’Node b or eNode B ,an Enode B is the Base Transceiver Station(BTS) for LTE. Radio resource management is completed by the eNodeB * eNode B are connected to theEvolved Packet Core (EPC), using a Mobility Management Entity(MME) for controlplane signalling.and a Serving Gateway for user plane data * The Serving Gateway is connectedto a Packet Data Network (PDN)Gateway for connectivity to external networksincluding the public internet Figure 1 - Long Term Evolution (LTE) architecture |
* Figure 2 illustrates a radioaccess network which includes macro, micro, pico and femto BTS, as well asrelays and repeaters. This type of radio access network is known as aheterogeneous network because it includes a range of different BTS types * The most common BTS type is themacro BTS, which 3GPP categorises as a ‘wide area BTS’. 3GPP categorises picoBTS as 'local area BTS’ and femto BTS as 'home BTS’. 3GPP does not specify aseparate power class for micro BTS but a wide area BTS with reduced transmitpower could be designed and used as a micro BTS. 3GPP also specifies separaterequirements for repeaters and relays * Macro BTS are characterised byhaving their antenna above roof-top level so their coverage area is relativelylarge Their transmit power is typically 20.40 or 60 Watts, and they normallyhave more than a single sector * Micro BTS are characterised byhaving their antenna below roof-top level so their coverage tends to be limitedby the neighbouring buildings. Their transmit power is typically 5 or 10 W, andthey often have only a single Sector * Pico B I S are designed to providecoverage and capacity across small areas. Their transmit power does not exceed 0.25 W so their antennaneeds to be close to the source of traffic * Femto BTS are intended for use athome, or in small offices. Their transmit power does not exceed 0.1W so theyneed to be used in areas where coverage from other BTS types is relativelyweak, e.g. indoors In contrast to other BTS types,the location of Femto BIS isno usually controlled by the network operator. End-users are free to placetheir Femto BTS Wherever they like. The network architecture for Femto BTS alsodiffers from other BTS types.Femto BTS are connected to the Evolved Packet Coreusing a Home eNode B Gateway(also known as a Femto Gateway). The connectionbetween the Femto BTS and Femto Gateway typically uses a home broadband connection, e.g. ADSL * Repeaters can be used to extendthe coverage of an existing BTS They re-transmit the uplink and downlinksignals without having to decode any of the content. Repeaters have one antennadirected towards the donor cell, and a second antenna directed towards thetarget coverage area. The target coverage area could be an indoor location sothe second antenna could be indoors * Relays also rely upon an RFconnection to a donor cell, but Relays differ from Repeater because Relays havetheir own cells and their own protocol stack, i.e. a Relay is similar to anormal BTS but without a fixed transport connection.Relay decode signals andmake radio resource management decisions Figure 2 - Heterogeneous network for LTE |
* The various BTS types typicallyshare the same channel bandwidth, so heterogeneous networks generate challengesin terms of co-channel interference and RF planning to achieve the intendedcoverage from each site. Traffic management can also be challenging when ClosedSubscriber Groups (CSG) are used to ensure that only authorised subscribers canuse certain BTS, e.g. only the family owning a Femto BTS can use that BTS. ThatBTS then appears as a source of interference to other subscribers 2.3 INTERFACES * Figure 3 illustrates the mostimportant interfaces for the radio access network * The air-interface connectionbetween the User Equipment (UE) and the eNode B is known as the U /u. The UEand eNode B make use of the U/u interface whenever they transmit or receiveacross the LTE air-interface * The X2 interface connects one eNodeB to another eNode B. This allows both signalling and data to be transferredbetween neighbouring eNode B ¡ thecontrol plane of the X2 (X2-CP) interlace allows signalling between eNode B ¡ theuser plane of the X2 (X2-UP) interface allows the transfer of application databetween eNode B * the S1 interface connects an eNodeB to the Evolved Packet Core (EPC). This allows both signalling and data to betransferred between the Evolved Packet Core (EPC) and Evolved UMTS Radio AccessNetwork (E-UTRAH) ¡ the controlPlane of S1 (Sl-MME)interface allows signalling with the MME ¡ the user planeof the S1(SI-UP) interface allows application data transfer through the Serv ing Gateway * Application Protocols have beenspecified to difine the signaling procedures and message types which can besent across the X2 and S1 interfaces,i.e. X2-AP and S1-AP
* Both the X2 and S1 interfaces arebased upon IP * Figure 3 illustrates a logicalrepresentation of the interfaces within E-UTRAN. In practice, the X2 and S1 arelikely to use a single physical connection at the eNode B, i.e. a singleEthernet cable can be used for both the X2 and S1 interfaces * Figure 4 illustrates an examplephysical representation of the X2 and S1 interfaces. The eNode B are connectedto an IP backhaul transport network using a single Ethernet cable. This cabletransfers information for both the X2 and S1 interfaces * In the case of the X2 interface,the IP routers within the transport network receive data from one eNode B anddirect it towards another eNode B. In the case of the S1 interface, the IProuters provide connectivity between the eNode B and the Evolved Packet Core * The Ethernet conneetion betweenthe eNode B and transport network could be based upon either an electrical oroptical Gigabit Ethernet cable * IP Quality of Service (QoS) can beused to differentiate and prioritise packets transferred across the IP backhaul * Timing over Packct (ToP) can beused to provide the eNode B with synchronisation information. ToP is specifiedwithin IEEE 1588. Alternatively, Global Positioning System (GPS) satellites canbe used, or a synchronisation signal can he provided by a co-sited BTS
Figure 4 - Key interfaces for LTE (physical representation) |
* A more complete set of interfacesassociated with LTE and the Evolved Packet Core is shown in Figure 5. Only asingle eNode B is shown in this figure so the X2 interface does not appear. Thecontrol plane of the S1 interface is shown as the S1-MME, while the user planeis shown as the S1-UP * The S11 interface connects the MME.to the Serving Gateway. This allows signalling information for mobility andbearer management to be transferred. Application data does not use the S11interface * The S5 interface connects theServing Gateway to the Packet Data Network (PDN) Gateway. Both control planesignalling and user plane data use the S5 interface. The PDN Gateway providesconnectivity to the set of IP services so the S5 represents the main connectionfor application data across the Evolved Packet Core * The S8 interface is similar to theS5 interlace but it terminates at a PDN Gateway belonging to a different PLMN.This interface is used by end-users who are roaming away from their home PLMN * The S6a interface connects the MMEto the Home Subscriber Server (HSS). The HSS hosts a database containingsubscription related information for the ******lation of end-users. The HSSrepresents an evolution of the Home Location Register (HLR) used by earliernetwork architectures * The S13 interface connects the MMEto the Equipment Identity Register (EIR). The EIR stores the InternationalMobile Equipment Identities (IMEI) of the end-user devices used within thenetwork. These IMF,I can be ‘white listed’, ‘grey listed’ or ‘black listed’ tocontrol access to the network * The Gx interface connects thePolicy and Charging Enforcement Function (PCEF) within the PDN Gateway to thePolicy and Charging Rules Function (PCRF). The PCRF provides QoS and charginginformation to the PDN Gateway. The Gx interface is also known as the S7interface in some references * The SGi interface providesconnectivity between the PDN Gateway and a packet data network. The packet datanetwork could be an external network (either public or private), or couldbelong to the operator. The SGi interface corresponds to the Gi interface inearlier network architectures * The S3 interface allows thetransfer of control plane signalling between the MME and an SGSN. The SGSNcould belong to either a UMTS or GPRS network. The main purpose of thesignalling is to allow mobility between the various access technologies * The S4 interface allows thetransfer of application data between the Serving Gateway and SGSN when a‘Direct Tunnel’ is not established between the RNC and Serving Gateway. Thisinterface may be used when a UE roams from the LTE network across to a UMTSnetwork * The S2a interface providesconnectivity between the PDN Gateway and a non-3GPP access technology. Figure 5illustrates the non- 3GPP technology as a wireless LAN. WiMax is a non-3GPPacccss technology which could be connected using the S2a interface Figure 5 - Additional interfaces for LTE and the Evolved Packet Core |
* The SI2 interface allows thetransfer of application data between the Serving Gateway and RNC when a ‘DirectI unnel is established. The S4 interface represents the alternative when a‘Direct Tunnel’ is not established. Both the S12 and S4 interfaces areapplicable when a UE roams from the LTE network across to a UMTS network * 3GPP References: TS 36.410, TS36.420, TS 23.002, TS 23.402
2.4 CHANNEL BANDWIDTHS * 3GPP has specified a set of 6channel bandwidths, ranging from 1.4 MHz to 20 MHz. These are presented inTable 1
| | | | | | | | | Number of Resource Blocks | | | | | | | | Number of Subcarriers | | | | | | | | Uplink Subcarrier Bandwidth (MHz) | | | | | | | | Downlink Subcarrier Bandw idth (MHz) | | | | | | |
Table 1 - Channel bandwidths for LTE. * A Resource Block represents thebasic unit of resource for the LTE air-interface. The eNode B schedulerallocates Resource Blocks to UE when allowing data transfer * The subcarriers belong to theOrthogonal Frequency Division Multiple Access (OFDMA) technology in thedownlink, and the Single Carrier Frequency Division Multiple Access (SC-FDMA)technology in the uplink * There are 12 subcarriers perResource Block so the number of subcarriers equals 12 x number of Resource Blocks * Each subcarrier occupies 15 kHz sothe total subcarrier bandwidth equals 15 kHz x number of subcarricrs * The downlink subcarrier bandwidthincludes an additional 15 kHz to accommodate a null subcarrier at the center ofall other subcarriers. The null subcarrier provides 15 kHz of empty spectrumwithin which nothing is transmitted * The total subcarrier bandwidth isless than the channel bandwidth to allow for the roll-offofemissions and toprovide some guard band * The larger channel bandwidthsprovide support for the higher throughputs. Smaller channel bandwidths providesupport for lower throughputs but are easier to accommodate within existingspectrum allocations * 3GPP also specifies a subcarrierspacing of 7.5 kHz (in addition to the subcarrier spacing of 15 kHz). T hesubcarrier spacing of 7.5 kHz is only used in cells which are dedicated toMultimedia Broadcast Multicast Services (MBMS). There are 24 rather than 12subcarriers per Resource Block when using the 7.5 kHz subcarrier spacing so thetotal bandwidth of a Resource Block remains the same * LTE Advanced provides support forCarrier Aggregation which allows multiple ‘Component Carriers’ to be used inparallel. This effectively increases the channel bandwidth to the sum of theindividual Component Carriers 2.5 FREQUENCY AND TIME DIVISION DUPLEXING * LTE has been specified to supportboth Frequency Division Duplexing (FDD) and Time Division Duplexing (TDD) Theconcepts of FDD and TDD are illustrated in Figure 6 Figure 6 - FDD and TDD concepts |
* FDD is based upon using twoseparate RF carriers for uplink and downlink transmission, i.e. the UEtransmits using one RF carrier (uplink), while the BTS transmits using adifferent RF carrier (downlink) * TDD is based upon using the sameRF carrier for both the uplink and downlink transmissions. The UE and BTScannot transmit simultaneously in the case of TDD because they share the sameRF carrier * FDD uses frame structure ‘type 1 \whereas TDD uses frame structure ‘type 2’. These frame structures are presentedin section 3.2 * TDD is attractive for systemswhere the data transfer is highly asymmetric because the ratio between theuplink and downlink transmissions can be adjusted appropriately and the RFcarrier remains fully utilised. In the case of FDD, one of the RF carriers wouldbe under utilised when the data transfer is highly asymmetric * TDD devices benefit from notrequiring a duplexer. This helps to reduce the cost of the device. A duplexeris required by FDD devices to extract the uplink signal from the antenna, whileat the same time inserting the downlink signal into the antenna. Duplexers tendto increase the receiver noise figure in the receive direction and generate anadditional loss in the transmit direction * FDD is attractive for systemswhere the requirement for uplink and downlink capacity is relatively symmetric.FDD can offer higher throughputs because data transfer can be continuous inboth directions. The capacity associated with a pair of FDD carriers is greaterthan the capacity associated with a single TDD carrier, but a greater quantityof spectrum is required * FDD can be simpler to deploy interms of synchronisation requirements. In general, it is not necessary forneighbouring FDD BTS to be time synchronised. Neighbouring TDD BTS require timesynchronisation to limit levels of interference between uplink and downlinktransmissions * LTE also supports a thirdduplexing technology known as half duplex FDD. The concept of half duplex FDDis illustrated in Figure 7 Figure 7 - Half duplex FDD concept |
* In the case of half duplex FDD,the BTS is able to transmit and receive simultaneously, but the UE is not ableto transmit and receive simultaneously. Both uplink and downlink RF carrierscan be fully utilised by time multiplexing different UE * Half duplex FDD uses framestructure ‘type 1 i.e. the same frame structure as FDD. This frame structure ispresented in section 3.2 * The BTS scheduler is responsiblefor providing half duplex operation by ensuring that UE do not need to transmituplink data at the same time as receiving downlink data. This has to accountfor the requirements to send and receive acknowledgements after data has beentransferred * Similar to TDD, half duplex FDDcan be an attractive solution because it avoids the requirement for a duplexerwithin the UE so helps to reduce the cost of devices. This argument isespecially valid for operating bands which have small duplex separations(frequency separation between the uplink and downlink RF carriers). Duplexerdesign becomes more challenging and more expensive when the uplink and downlinkoperating bands are relatively close to each other
| 
发表于 2014-11-4 16:41:46
