Sei sulla pagina 1di 5

The soft-cell approach

Heterogeneous network deployments in LTE


Complementing high-power macro nodes with lower-power ones is an attractive means of meeting the predicted requirements for higher data rates and additional capacity.
S T E FA N PA R K VA L L , E R I K DA H L M A N, GE ORGE JNGR E N, S A R A L A N D S T RM A N D L A R S L I N DB OM

LTE is rapidly emerging as the worlds most dominant 4G technology, taking mobile broadband to unprecedented performance levels. To meet expectations and predictions for even higher data rates and traffic capacity beyond what is available in current LTE networks a densified infrastructure is needed 1. In scenarios where users are highly clustered, using multiple low-output power sites to complement a macro cell providing basic coverage is an attractive solution as illustrated in Figure 1.
This strategy results in a heterogeneousnetwork deployment with two cell layers. The principle can be extended to more than two layers and the concept of multiple layers, is in itself not new; hierarchical cell structures have been considered since the mid-1990s but, at that time, the discussion applied to mobile technologies primarily offering low-rate voice services.

In this article, the discussion focuses on radio-interface solutions, standardized by 3GPP, to enhance the performance of heterogeneous-network deployments (or heterogeneous deployments) with all nodes operating on the same frequency. Traditionally, a terminal connects to the node from which the downlink signal strength is the strongest. In Figure 1, the solid orange areas are those in which the signal from the corresponding pico node is the strongest. Users in these zones connect to the appropriate low-power node. Due to the difference in transmission power between the pico nodes and the overlying macro node, this strategy does not necessarily result in the terminal connecting to the node to which it has the lowest path loss as illustrated in Figure 2. It is, therefore, not the best node-selection strategy for achieving high uplink data rates. The uptake area of a low-power node can be expanded without increasing the output power of the node by adding an offset to the received downlink signal strength in the cell-selection

mechanism. In practice, some additional factors such as backhaul capacity should also be included in the cell selection process. Increasing the uptake area of a node is sometimes referred to as range expansion. The advantages of this technique are:
enhanced uplink data rates by at least partially taking uplink path loss into account when associating terminals with a low-power node; increased capacity receiving downlink traffic from the low-power node even if the received signal strength from the macro is higher allows for the reuse of transmission resources across low-power nodes; and improved robustness enlarging the coverage area of a low-power node can reduce its sensitivity to ideal placement in a traffic hotspot.

BOX A 

Terms and abbreviations


3rd Generation Partnership Project 4th Generation mobile wireless standards almost blank subframe broadcast channel carrier aggregation cell-specific reference signal demodulation-specific reference signals inter-cell interference coordination Long Term Evolution PDCCH PDSCH PSS RE RRC RRU Rx SSS UL CoMP physical downlink control channel physical downlink shared channel primary synchronization signal range expansion Radio Resource Control remote radio unit radio receiver secondary synchronization signal uplink coordinated multipoint reception

3GPP 4G ABS BCH CA CRS DM-RS ICIC LTE

A heterogeneous deployment, with a modest range expansion somewhere in the region of 3-4dB, is already possible in the rst release of LTE, Rel-8. The benets gained from range expansion are highly dependent on the individual scenario and, in many cases, modest range expansion is best. Nevertheless, 3GPP has recently discussed the applicability of excessive range expansion with cellselection offsets up to 9dB. Such deployments are particularly problematic, as a terminal in the range-expansion zone (the striped area shown in Figure 2) may experience very low downlink signalto-interference ratio due to the signicant difference in output power of the nodes. Specically, downlink control signaling in the range expansion zone which is essential for the low-power node to control transmission activity poses a problem. Transmission of the

E R I C S S O N R E V I E W 2 2011

data part is less challenging as Rel-8 supports methods for ensuring non-overlapping transmissions in the frequency domain from the macro and the lowpower node using inter-cell interference coordination (ICIC)2. This article discusses two different approaches to heterogeneous deployment resource partitioning and softcell schemes both of which provide support for excessive range expansion. Resource partitioning By restricting macro-cell transmissions from using the same time-frequency resources as the low-power node, control signaling from the low-power node to the terminal can be protected. Resource partitioning can be implemented in either the frequency domain, by using support for carrier aggregation (Rel-10), or in the time domain, by relying on almost blank subframes (ABSs), a feature that will be fully supported in LTE Rel-11 (see Figure 3). Frequency-domain partitioning This method protects downlink controlsignaling from the low-power node in the range-expansion zone by placing control signaling from the macro and low-power nodes on separate carriers as illustrated in Figure 3. Assuming transmissions from low-power nodes are time synchronized with the overlying macro, the control signaling on carrier f 2 in the range-expansion zone will not be subject to major interference from the macro node. At the same time, through the use of carrier aggregation, data transmissions can still benet from the full bandwidth of both carriers. The Rel-8 ICIC mechanism can be used to coordinate use of data resources. Regardless of the extent of range expansion, frequency-domain partitioning is a natural choice to support heterogeneous deployments for operators who already rely on carrier aggregation (CA) to exploit fragmented spectrum; and who have a reasonable number of subscribers using CA-capable terminals in their networks. Time-domain partitioning This method protects the downlink control-signaling from the low-power node by reducing macro transmission activity in certain subframes

FIGURE 1 

Heterogeneous deployment

FIGURE 2 

Range expansion
Low-power node
Rx p (pat owe

h los

Range expansion zone

s) -1

Macro cell

Downlink-signal-strength cell border Path-loss-based cell border

FIGURE 3 

Frequency-domain and time-domain partitioning

Carrier aggregation (CA) Almost blank subframes (ABSs)

f1 f2 f

E R I C S S O N R E V I E W 2 2011

The soft-cell approach

FIGURE 4A 

Independent cells
PSSA, /SSSA, BCHA, CRSA

PSSB, /SSSB, BCHB, CRSB

PSSC, /SSSC, BCHC, CRSC

Cell B

Cell A

Cell C

FIGURE 4B 

Soft cell

PSSA, /SSSA, BCHA, CRSA

which is illustrated in the bottom part of Figure 3. The low-power node is provided with data about the set of protected subframes over the X2 interface and can use this information when scheduling users who are in the rangeexpansion zone. For backward compatibility, the macro node must transmit certain signals, most notably cell-specic reference signals (CRSs) and synchronization signals (PSSs/SSSs), in downlink subframe in the same way as in Rel-8. The protected subframes are, as a result, not completely blank but they are almost blank. Terminals need to apply interference suppression to receive control signaling from the low-power node. Time-domain partitioning can thus be viewed as a terminal-centric approach to achieving excessive range expansion. Support for time-domain partitioning for excessive range expansion is incomplete in Rel-10; X2 and RRC signaling are included, whereas interference-suppression receivers are still under discussion for Rel-11. The main argument for implementing timedomain partitioning is to enable support for excessive range expansion for those operators that do not want to rely on carrier aggregation. Soft-cell schemes In both frequency-domain and timedomain partitioning schemes, the lowpower nodes create separate cells, each of which has an individual cell identity that differs from that of the macro cell. As a consequence, each pico node transmits unique system information and synchronization signals, including reference signals as illustrated in Figure4A. In an alternative approach known as shared cell or soft cell, lowpower nodes can be part of the macro cell without creating independent cells as illustrated in Figure 4B. The distinction between cell and transmission points is an important aspect of the soft-cell approach. Each cell has a unique cell identity from which the CRS is derived. With the cell-identity information, a terminal can derive the CRS structure of the cell and obtain the system information it needs to access the network. A transmission point, on the other hand, is simply one or more collocated antennae from

Cell A

FIGURE 5 

Heterogeneous deployment using a soft-cell scheme

Same PSS/SSS, BCH, CRS

2
Data (PDSCH) Control (PDCCH)

E R I C S S O N R E V I E W 2 2011

which a terminal can receive data transmissions. Note that the sectors of a site constitute separate points. Traditionally, each cell has one transmission point from which the CRS, as well as all data transmissions, are sent. In Rel-10, however, demodulation-specic reference signals (DM-RSs) were introduced. Unlike CRSs, these signals are subject to the same pre-coding as the associated data and are transmitted only when a corresponding data transmission is detected. The terminal can deduce the channel needed for demodulation based on the fact that both the DM-RS and data are transmitted in a similar manner. This implies that DM-RS-based data transmission to a terminal does not have to be sent from the transmission point used for CRSbased information, and that time-frequency resources for data can be reused at different transmission points. In Figure 5, data is transmitted to terminal 1 from the low-power node farthest to the left. Since the associated DM-RS is transmitted from the same transmission point as the data, the terminal does not need to know which point is used for data transmission to achieve area-splitting gains the reuse of time-frequency resources for data transmission across multiple low-power nodes within the same macro cell. The control information required in Rel-10 is based on CRS and needs to be transmitted from, at least, the macro site. In many cases, this results in data and associated control signaling originating from different transmission points. This is transparent to the terminal; it needs to match the reference signal with the corresponding data signal. The identity of the transmission point, on the other hand, is irrelevant. Figure 5 shows different ways to transmit control information. The case for terminal 1 where control signaling originates only from the macro site has already been described. This method results in reduced network energy consumption, because the low-power node is active only when there is data to transmit. For terminal 2, identical CRS and control signals can be transmitted from the macro and the low-power node. As the same signal is transmitted from both nodes, the terminal will interpret them

FIGURE 6 

Comparison of different approaches to heterogeneous deployments


Any backhaul Medium RE Separate cell Excessive RE Low-latency backhaul (allows for UL CoMP) Any RE CRS-based data and control DM-RS-based data CRS-based control DM-RS-based data and control Resource partitioning Carrier aggregation Rel 8 Almost blank subframes

Soft cell

Rel-8

Rel-10

Rel-10+

as a single composite node. This method results in an improved signal-to-noise ratio for control signaling through an over-the-air combination of transmissions from both the macro and the lowpower nodes. For power-control purposes, LTE terminals estimate the uplink path loss from the strength of the received CRS signal. Consequently, the case illustrated by terminal 2 can sometimes result in more accurate uplink power control at least for Rel-8/9/10 terminals. A minor update to a future release of the LTE standard is currently under discussion in 3GPP that will provide the same uplink power-control accuracy while allowing for greater energy efciency in network operations. To further improve the performance of the soft-cell scheme, enhancements to support DM-RS-based control signaling are likely to be included in LTE Rel11. This will provide area-splitting gains for control signaling, in contrast to the CRS-based control signaling in Rel-10 and previous releases. Terminals from previous releases that do not support DM-RS-based transmission can still operate in a soft-cell scheme without any area-splitting gains. Data transmission to such terminals is CRS-based and is handled in the same way as control signaling. These terminals will benet from low-power nodes because of the improved signalto-noise ratio. A soft-cell scheme can be deployed

by connecting one or several RRUs and the macro site to the same main unit. In practice, this link should use high-speed microwave or optical ber, as it requires low-latency and a fairly high-capacity connection for tight coupling between the macro and low-power nodes where control and data signaling originate from different transmission points. However, with the availability of DM-RSbased control signaling, backhaul requirements will be relaxed as both the data and control signaling can originate from the same transmission point. The centralization of processing provides benets in uplink performance, which is often signicant enough to justify using RRUs with centralized processing irrespective of range expansion. Any combination of points not necessarily those used for downlink transmission to a terminal can be used to receive transmissions from a terminal. By combining the signals from different antennae in a constructive manner at the central processing node a method known as uplink softer handover a signicant improvement in uplink data rates can be achieved. In addition to avoiding much of interfering CRS transmissions heterogeneous deployments that use soft cells can provide greater mobility robustness than deployments with separate cells. This is important, especially when moving from a low-power node to the macro. In separate cell deployment, a handover procedure is required to
E R I C S S O N R E V I E W 2 2011

The soft-cell approach

switch serving cells. If, during the time it takes to perform the handover procedure, the terminal has moved too far into the macro area, it may drop the downlink connection from the lowpower node before handover is complete leading to a radio-link failure. In softcell deployment, the transmission point that should be used for downlink transmission can be changed rapidly without a handover procedure thus reducing the probability of dropped connections. Conclusion A heterogeneous-network deployment is a favorable means of meeting future data-rate and capacity demands. In many cases, the support provided in Rel8 is sufcient. This article has provided an overview of various schemes, summarized in Figure 6, including carrier aggregation, almost blank subframes and soft cell with a focus on the latter. The choice of scheme depends on the scenario, although the network-centric soft-cell approach provides many benets without the requirement for not-yetstandardized terminal functionality.

Stefan Parkvall
is currently a principal researcher at Ericsson Research, with a focus on future radio access. Parkvall has been heavily involved in the development of HSPA, LTE and LTE-Advanced radio access. He is a senior member of IEEE and co-author of the book 3G Evolution HSPA and LTE for Mobile Broadband and 4G LTE/LTE-Advanced for Mobile Broadband. In 2009, he was a corecipient of Stora Teknikpriset (Swedens major technology award) for his work on HSPA. In 1996, he received a Ph.D. in electrical engineering from the Royal Institute of Technology (KTH) Stockholm. He was previously an assistant professor in communication theory at KTH and a visiting researcher at University of California, San Diego, in the US.

George Jngren
joined Ericsson Research in 2005 and is a master researcher in the area of radio-access technologies. His current focus is on research and standardization of multiantenna and coordinated multi point (CoMP) techniques for LTE. During his early years at Ericsson he was part of the development of the MIMO HSDPA test-bed. He holds an M.Sc. (1998 ) and Ph.D. (2003) in electrical engineering from the Royal Institute of Technology (KTH), Stockholm, Sweden. In 1997, he was elected Teacher of the Year in electrical engineering at KTH.

Sara Landstrm
is a senior researcher at Ericsson Research in Lule, Sweden. Her research area is Wireless Access Networks and her current focus is heterogeneous networks. She has also been involved in evaluating IMT-Advanced candidate technologies and service-oriented research. She joined Ericsson in 2008 after receiving her Ph.D. in computer networking from Lule University of Technology, Sweden.

Erik Dahlman
joined Ericsson Research in 1993 and is a senior expert in the area of radio access technologies. He has been deeply involved in the development and standardization of 3G radio access technologies (WCDMA/HSPA) as well as LTE and its evolution. He is part of the Ericsson Research management team working on long-term radio access strategies. He is also co-author of the book 3G Evolution HSPA and LTE for Mobile Broadband and 4G LTE/LTE-Advanced for Mobile Broadband. In 2009, together with Stefan Parkvall, he received the Stora Teknikpriset award in 2009 for contributions to the standardization of HSPA. He holds a Ph.D. from KTH in Stockholm.

Lars Lindbom
currently holds a position as systems manager at Ericsson Business Unit Networks, where he works on concepts and standards for future radio access, including standardization related to heterogeneous networks for 3GPP. He has a Ph.D. in signal processing from Uppsala University.

 References

1. S. Landstrm, A. Furuskr, K. Johansson, L. Falconetti, and F. Kronestedt, Heterogeneous networks (hetnets) an approach to increasing cellular capacity and coverage, Ericsson Review, No 1, 2011. 2. E. Dahlman, S. Parkvall, and J. Skld, 4G: LTE/LTE-Advanced for Mobile Broadband, Elsevier, 2011.
E R I C S S O N R E V I E W 2 2011

Potrebbero piacerti anche