Citation Context

...area is far lower than during a normal subframe, it is not completely zero. Usually the remainingssignal power is around 9% of the power in normal transmission mode, assuming a 2x2 MIMO configuration =-=[36]-=-. Only advanced UE terminals with the ability of cancelling CRS interference willsexperience nearly zero interference during ABS. This new feature included in 3GPP Release 11 isscalled CRS interferenc...

Citation Context

...ease in capacity would have been achieved by adding more macro nodes insthe network. However, the high costs and the space needed for such an approach represent an important problem for the operators =-=[2]-=-. Moreover, spectral efficiency per link is reaching theoreticalslimits [3].sAt the sight of this, LTE-Advanced (LTE-A) in conjunction with Heterogeneous Networkss(HetNets) are seen as the keys for fu...

Citation Context

...crease the capacity in zones with very high traffic volume (usually called hotspots) [4]. Therefore,sHeterogeneous Networks are envisioned as the major performance improvement enablers of LTEAdvanced =-=[28]-=-. A possible HetNet deployment is shown in Figure 2.10.sCoverage area of macro eNB Macro eNB UE Signal ~ 18 ~sRadio Access Technology Sections~ 19 ~sFigure 2.10: Heterogeneous Network Deployment using...

Citation Context

... IX ~sRadio Access Technology SectionsList of FiguressFigure 1.1: Site-specific operator scenario (left) and 3GPP scenario (right) network layout .............. 3sFigure 2.1: LTE Network Architecture =-=[19]-=- ....................................................................................... 9sFigure 2.2: LTE Downlink Frame Structure [19] ..................................................................

Citation Context

...ase 8. For that reason, 3GPP also defined its own requirements for LTE-Advanced that, in some areas, exceed the ones issued by thesITU-R, especially in terms of peak and cell-edge spectral efficiency =-=[16]-=-. This ensures an incremental step of performance and capabilities between the successive releases. More detailed informationsabout LTE-A requirements can be found in [17].sThe performance improvement...

Citation Context

... measurement of the received signal strength. The serving cell selected is the one with theshighest signal. There are different signal strength indicators. In LTE the measurements carried onsare [19] =-=[29]-=- [30]:s• Reference Signal Received Power (RSRP), which measures for each cell the average of the power of the resource elements that carry cell-specific reference signalsswithin the considered measure...

Citation Context

...on expressed as the sum of log(...

Citation Context

.... Theslatter is a more advanced model that includes separate models for line-of-sight (LOS) and non-LOSs(NLOS) links, spatial correlated models for shadow fading, among other assumptions explained ins=-=[42]-=-. Regarding the antennas, a directional 3D antenna pattern with down-tilt is modeled for thesmacro cells, while small cells are simply equipped with omni-directional antennas.sFigure 3.1: Example of 3...

Citation Context

...ich gets higher speeds, marked the developmentsof mobile broadband and data oriented devices such as smartphones and USB modems. These devices are responsible for a fast growth in mobile data traffic =-=[1]-=-. In 2012, global mobile data trafficsgrew 70 percent and it is expected to increase 13-fold between 2012 and 2017. Moreover, the number of laptops and tablets mobile-connected is increasing exponenti...

Citation Context

...sically takes into account the past average throughput over all the configured carriers (for proportional fair scheduler case) when calculating thesscheduling metric (see equation (2.1)). As shown in =-=[50]-=-, this type of scheduling mechanism pro~ 65 ~sRadio Access Technology Sectionsvides a more fair resource sharing among the users especially in the cases where some users aresserved only by one cell, w...

Citation Context

... keeps backwards compatibility with older LTE terminals.sIn the specific case of LTE-Advanced, up to 5 CCs of 20 MHz can be used simultaneouslyswhich leads to a substantial increase in the data rates =-=[26]-=-. There are three possible carrier aggregation scenarios: contiguous aggregation of carriers in a single band, non-contiguous aggregation ofscarriers in a single band and non-contiguous aggregation of...

Citation Context

...figuration for a network, as well as a great amount of other studies tryingsdifferent techniques and algorithms to reach an optimal user association. Extensive information cansbe found in [5] [6] [7] =-=[8]-=-.sThe usage of Range Extension leads to interference problems as the carrier frequency issshared between the high-power and the low-power layers. A strong control and management of thesinter-layer int...

Citation Context

... macro and small cells, resulting in few users served by the small cells. It is desirable forsthe network to balance the load between the macro and pico layers to harvest the gain from cellssplitting =-=[31]-=-. The most commonly used technique is the Range Extension offset [4] . Besides, researchers have also been proposing other algorithms for selecting the serving cell [32] [33] [34] [5]s[6] .sRange Exte...

Citation Context

................. 14sFigure 2.7: Layer 2 RRM Functionalities [22] ................................................................................. 15sFigure 2.8: RRM algorithms with Carrier Aggregation =-=[13]-=- ............................................................ 17sFigure 2.9: Traditional Network Deployment .................................................................................. 18sFigure...

Citation Context

.... Moreover, when outer loop link adaptation is used, theslink adaptation unit can also take information from the HARQ acknowledgements from past transmissions in order to obtain more accurate results =-=[23]-=-.sReporting a CQI for every PRB will lead to excessive uplink signaling. For this reason, ansLTE terminal can be configured to report information on specific sub-bands. It is important to mention that...

Citation Context

...tion about street and building sizes and positions.sFor outdoor areas, which make up for approximately 60% of the analyzed region, thespathloss prediction was calculated using the dominant path model =-=[45]-=- [46]; for indoor locations, ansoutdoor-to-indoor penetration loss of 20 dB plus 0.6 dB per indoor meter is added to the highestsreceived power among the external walls. Pathloss maps were generated a...

Citation Context

........................................ 13sFigure 2.6: Overview of the RRM functionalities in each layer [19] ................................................ 14sFigure 2.7: Layer 2 RRM Functionalities =-=[22]-=- ................................................................................. 15sFigure 2.8: RRM algorithms with Carrier Aggregation [13] ............................................................

Citation Context

... gain from cellssplitting [31]. The most commonly used technique is the Range Extension offset [4] . Besides, researchers have also been proposing other algorithms for selecting the serving cell [32] =-=[33]-=- [34] [5]s[6] .sRange ExtensionsA positive bias factor is added to the RSRP level of each pico cell, thus expanding theirscoverage area. This offset is called Range Extension (RE) [4]. A typical deplo...

Citation Context

...ifferences between the areas, especially in terms of macro and pico cell density, user distribution, etc. These and other factors affectsthe potential benefit of the multi-cell cooperation techniques =-=[48]-=- [49]. For example, larger gainssare likely to be obtained in areas with high small cell density (e.g. area 4); whereas, no benefit fromseICIC/inter-site CA is expected in area 5, which is a macro-onl...

Citation Context

...t RE configuration for a network, as well as a great amount of other studies tryingsdifferent techniques and algorithms to reach an optimal user association. Extensive information cansbe found in [5] =-=[6]-=- [7] [8].sThe usage of Range Extension leads to interference problems as the carrier frequency issshared between the high-power and the low-power layers. A strong control and management of thesinter-l...

Citation Context

...yer with very bad channel quality does not bring any gains but costs the additional resource consumption in the network, as explained previously.s~ 30 ~sRadio Access Technology SectionsRSRQ Threshold =-=[38]-=- [39]sFor UEs configured with inter-site CA, the assumption is that users have their Primary Cells(PCell) configured on the best macro cell, with the option of also having a small cell configured asst...

Citation Context

... configuration for a network, as well as a great amount of other studies tryingsdifferent techniques and algorithms to reach an optimal user association. Extensive information cansbe found in [5] [6] =-=[7]-=- [8].sThe usage of Range Extension leads to interference problems as the carrier frequency issshared between the high-power and the low-power layers. A strong control and management of thesinter-layer...

Citation Context

...leases. More detailed informationsabout LTE-A requirements can be found in [17].sThe performance improvement brought by LTE-A is mainly achieved thanks to the introduction of new features. As seen in =-=[18]-=-, some of these are:s• Intra and Inter-band Carrier aggregation (CA)s• Enhanced MIMOs• Coordinated multipoint transmission and reception (CoMP)sApart from these, LTE-A also introduces the eICIC functi...

Citation Context

... protocol stack. In the uplink direction, the UE physical layersmust incorporate N parallels SC-FDMA transmitters in order to support non-contiguous aggregationsof carriers. However, as shown in [16] =-=[27]-=-, the use of multiple carriers in the uplink has severalsimplications in the UE transmission power (power backoff issues); therefore, allocating more CCssto a UE in UL direction might not always resul...

Citation Context

...t the users in the RE area. For this purpose, the techniquesknown as enhanced Inter-Cell Interference Coordination (eICIC) has been studied and showedspromising results. Some of these studies are [3] =-=[9]-=-. The authors in [10] [11] introduce the FastsMuting Adaptation, a dynamic version of eICIC.sIn the case of a dedicated carrier deployment, the inter-layer interference is not a problemsas macro and l...

Citation Context

...urement of the received signal strength. The serving cell selected is the one with theshighest signal. There are different signal strength indicators. In LTE the measurements carried onsare [19] [29] =-=[30]-=-:s• Reference Signal Received Power (RSRP), which measures for each cell the average of the power of the resource elements that carry cell-specific reference signalsswithin the considered measurement ...

Citation Context

...about street and building sizes and positions.sFor outdoor areas, which make up for approximately 60% of the analyzed region, thespathloss prediction was calculated using the dominant path model [45] =-=[46]-=-; for indoor locations, ansoutdoor-to-indoor penetration loss of 20 dB plus 0.6 dB per indoor meter is added to the highestsreceived power among the external walls. Pathloss maps were generated at a h...

Citation Context

...ern must be adjusted adaptively according to the traffic and interference conditions to getsthe best overall performance in the network, as using ABS has a negative effect in the macro layerscapacity =-=[3]-=-. However, most studies have been focused in a slow muting adaptation, assuming ans~ 23 ~sRadio Access Technology Sectionsupdate in the ABS muting pattern each few seconds, due to the fact that this u...

Citation Context

...far. The idea is allowing somesuser equipments (UEs) to simultaneously connect to a macro cell and a small cell, thus benefitingsfrom larger transmission bandwidth from the two layers. The references =-=[12]-=- and [13] will enlightensthe reader about this technique.s1.2 Problem Statement and Objective of the StudysThere is extensive literature concerning the study of HetNets performance in generic 3GPPssce...

Citation Context

...t the gain from cellssplitting [31]. The most commonly used technique is the Range Extension offset [4] . Besides, researchers have also been proposing other algorithms for selecting the serving cell =-=[32]-=- [33] [34] [5]s[6] .sRange ExtensionsA positive bias factor is added to the RSRP level of each pico cell, thus expanding theirscoverage area. This offset is called Range Extension (RE) [4]. A typical ...

Citation Context

...improve the spectral efficiency per unit area, as the low-power cells makespossible to remove coverage holes in the macro-only network and increase the capacity in zonesswith very high traffic volume =-=[4]-=-. The main characteristic of HetNets is the great disparity betweensthe transmit power used by the high-power and the low-power nodes [2]. This makes necessary thesusage of interference management tec...

Citation Context

...quiress~ 22 ~sRadio Access Technology Sectionsexperienced human knowledge in order to get the best settings for optimizing the network performance.sFor the interested reader, some examples of studies =-=[5]-=- [6] which implement algorithms forsoptimal cell selection are briefly explained in Appendix B. In said appendix there is also a comparative summary table of techniques for optimal cell selection.sIn ...

Citation Context

...ith very bad channel quality does not bring any gains but costs the additional resource consumption in the network, as explained previously.s~ 30 ~sRadio Access Technology SectionsRSRQ Threshold [38] =-=[39]-=-sFor UEs configured with inter-site CA, the assumption is that users have their Primary Cells(PCell) configured on the best macro cell, with the option of also having a small cell configured asstheirs...

Citation Context

...nally placed in feasible locationsssuch as along streets or adjacent to buildings. More information about the small cell placement algorithm and the deployment criterion can be found in [43] and [37] =-=[44]-=-, respectively.s3.2.2 Traffic DistributionsFigure 3.4 shows the spatial traffic distribution in the operator-specific case. This trafficsdistribution is provided by the operator and corresponds to tra...

Citation Context

...y, a closed loop 2x2 single-user MIMO with pre-coding and rank adaptation issassumed for each link and the UE receiver type is Minimum Mean Square Error (MMSE)- Interference Rejection Combining (IRC) =-=[47]-=-. The UEs are considered to be LTE-Advanced compliant meaning that CQI measurement restrictions and CRS IC cancellation are assumed. Table 3.3scontains an overview of the general simulation assumption...

Citation Context

...re 2.11: HetNet topology using RE in the pico layer ................................................................. 21sFigure 2.12: Subframe classification and UE scheduling for slow ABS adaptation =-=[10]-=- .................. 24sFigure 2.13: Subframe classification and UE scheduling for fast ABS adaptation [10] ................... 25sFigure 2.14: Example of X2 signaling for distributed coordinated adapt...

Citation Context

...depend on the selected set of PRBs.sLTE is highly standardized by 3GPP; however, some of the network algorithms includingslink adaptation, power control and packet scheduling are vendor-specific [19] =-=[24]-=-.sIn this work, one of several packet scheduling implementation techniques is explained. Assproposed in [23], it is a two step algorithm: first, a time-domain packet scheduler (TDPS) selects upsto N p...

Citation Context

... from cellssplitting [31]. The most commonly used technique is the Range Extension offset [4] . Besides, researchers have also been proposing other algorithms for selecting the serving cell [32] [33] =-=[34]-=- [5]s[6] .sRange ExtensionsA positive bias factor is added to the RSRP level of each pico cell, thus expanding theirscoverage area. This offset is called Range Extension (RE) [4]. A typical deployment...

Citation Context

...e RE offsets without any interference management because that can increase the occurrence of Radio Link Failures (RLF) due tosthe poor channel quality experienced by users in the range extension area =-=[35]-=-.s• It is not straightforward to obtain the optimal settings of the RE for different networks as it depends on several things like the load of the network, the number ofsmacro and pico cells, the posi...

Citation Context

...ferent. It is usually assumed that the macro layer is deployed at a lower frequency than the pico layer to ensure a widescoverage area for the macro layer, as pathloss is smaller at lower frequencies =-=[37]-=-.sCA window Macro eNB Pico eNB UE served by macro cell only UE served by pico cell only UE in CA mode: receiving data from both cells RE pico UEsSignal ~ 29 ~sRadio Access Technology SectionsIn a dedi...

Citation Context

... cells were finally placed in feasible locationsssuch as along streets or adjacent to buildings. More information about the small cell placement algorithm and the deployment criterion can be found in =-=[43]-=- and [37] [44], respectively.s3.2.2 Traffic DistributionsFigure 3.4 shows the spatial traffic distribution in the operator-specific case. This trafficsdistribution is provided by the operator and corr...

Citation Context

...ences between the areas, especially in terms of macro and pico cell density, user distribution, etc. These and other factors affectsthe potential benefit of the multi-cell cooperation techniques [48] =-=[49]-=-. For example, larger gainssare likely to be obtained in areas with high small cell density (e.g. area 4); whereas, no benefit fromseICIC/inter-site CA is expected in area 5, which is a macro-only sec...