Antennas, how many do we need?

Antennas, how many do we need? Peter C.T. Song BEng(Hons) PhD CEng MIET SMIEE David Barker BEng(Hons) MSc CEng MIET MIEEE Quintel Technology Limited [email protected], [email protected]

Abstract: One of the key components interfacing between the radio hardware and the ‘Ether’, i.e. the air-interface for wireless communication is the antenna. With tremendous growth and demand for high speed high data rate wireless communication, more and more antennas, and antennas covering a multitude of frequency bands are required. This paper begins by giving an innovations overview of the evolution of the wireless standards and access technologies, driving antenna innovations from a conventional single narrow band component to today’s advances capable of multiple services, and cognitive capable antenna systems. The paper will focus on antenna developments in mobile communication systems, in particular cellular base station and cellular handset antennas. Key technological advancements will be presented. In conclusion, we will respond to the question of how many antennas do we really need.

1. Introduction The wireless delivery of information and services plays a crucial role to facilitate growth and prosperity. Today, it is one of the key measures to our modern day quality of life! Wireless communication has grown rapidly into today’s multitude of various high speed broadband radio standards. These wireless Wide Area Network (WAN) specifications such as cdma, GSM, WCDMA provides typically from a Macro to tomorrows Femto cellular coverage. From the network operator perspective, additional radio standards such as the Long Term Evolution (LTE) system are critical to managing the growing subscriber demands and data hungry services. In the same manner, wireless Personal Area Networks (PAN) such as Bluetooth, WiFi and UWB have also taken off promising short range high speed wire-free communication. This implies that mobile device manufacturers must not only operate within a multitude of

WAN radio standard but also equipped with PAN serviceability. Each wireless communication standard will require an antenna that resonates at its frequency of operation. With multiple standards and services, one could easily imagine a multitude of antennas required both at the base station site, and also on the mobile device. Antennas are often limited by size and bandwidth. This forms additional challenges to hoist more antennas onto an already packed cellular tower mast, or to pack the antennas into an ever shrinking mobile terminal platform. In section 2, we will look at different requirements for cellular base station and mobile handset antenna designs. In section 3, we look more specifically on the design of cellular base station antennas, and its evolution to a current state of the art operator antenna sharing technology. In section 4, we will see how the handset manufacturer has responded to these

advances with revolutionary solutions leading into cognitive antennas used in handsets. We will conclude to see how scalable antennas are compared to the development of these systems. 2. Mobile service evolution, the antenna Revolution Since the inception of 1G network back in the 1980s offering typically 2.4kbps, wireless communication has evolved rapidly into a multitude of different high speed broadband radio standards, addressing various application requirements. Today, improved 3G services such as HSPA+ could offer a single user up to 21Mbps peak rate theoretically, which is equivalent to the capacity 1,500 GSM voice subscribers! The loosely termed 4G networks promise many 10Mbps per user date rate. Interoperability is critically important, not only for consumer to seamlessly engage with different radio services, but most importantly, operators to optimise services on the different radio standards they provide. With rapidly diminishing cost of ownership for a mobile handset, subscriber traffic growth has been exponential over recent years, hungry for enhanced real time data services. This has not only pushed for more deployment of cell sites, but also the release of more spectrum and more spectral efficient communication standards. Polarisation diversity was also introduced during the 90’s in attempt to halve the number of antennas over the then more conventional spatial diversity deployment [1]. To date, network operators are slowly approaching their limit to accommodate the increased traffic, with finite spectral bandwidth. Operators often resort to cell splitting by adding further sites, or with more sectors at existing cell sites in areas of congestion. In densely populated areas where subscriber demand is often at its

highest, visual impact from a clutter of cellular antennas is often undesirable as shown in Fig 1, giving an additional hurdle to planning permission/zoning. The mobile handset market on the other hand has not only evolved technologically but into a fashion and iconic accessory. Mobile handsets are usually multi-radio compliant (e.g. GSM and WCDMA). On top of that, it is also armed with radio devices capable of personal area network communication such as WiFi, Bluetooth and Near-Field Communication (NFC) systems. Integration of these many services to a physically sizeable device becomes a critically important parameter.

Figure 1 Cluttering sights of cellular antenna masts over the years The radio spectrum, a natural resource, is fast becoming a scarcity. Studies have also shown that the allocated spectrum is not always fully utilized! [2] This prompted various exercises in spectrum re-farming and white space communication. The Cognitive radio concept is thus coined as an opportunistic radio, sensing out free spectrum to set up communication channels. Such ad-hoc and dynamic radio would then require a completely reconfigurable front end. More information can be found in [3].

It is clear that evolution in antenna design to provide a long term solution is required. Engineers are faced with challenges to realise multiband/wideband antennas, typically within a volumetric space available to the lowest operating band.

3. Advances in Cellular base station antenna 3.1 The cellular site Wireless operators face a multitude of factors that must be considered when deploying, maintaining, and expanding their networks. These factors, if not carefully considered, can have a negative impact on the system performance, such as coverage quality, co-channel interference, carrier-tointerference ratio and passive intermodulation distortion, given more and more spectrum bands are being used at site. As geographic distribution of network traffic varies, antennas on these cell sites often have to be optimised into a non ideal grid with different installation heights, beamwidths, and bearings to best manage the traffic distribution. Fig 2 shows an example of an optimised coverage tessellation produced from a network of cellsite antennas on non-regular grid with different locations, traffic distributions, heights, beam tilts and bearings.

Figure 2 Optimized cell-site coverage tessellation

. Tilting of antenna beam is therefore a crucial network optimisation technique for operators. It is often a balance between cochannel inter-site RF interference and service gaps between sites. Since each operator’s technology layers in a network are unique serving different traffic distributions and having different carrier to interference (C/I) requirements, it is fundamentally not ideal to share the same physical antenna due to their independent beam tilt optimization requirement. This results in more and more antennas being deployed to serve more and more access technologies. 3.2 The Antenna and challenges The Cellular Base Station (BS) antenna consists of typically an array of antenna elements packed linearly along a ground plane. Fig 3 shows the arrangement of a typical BS antenna site configuration, with electrical tilt function enabled via a motor and phase shifters.

Figure 3 Conventional dual slant polarisation base station antenna implemented with electrical tilt function via motor

In recent years, we have witnessed the technological advances where antenna engineers have packed as many array stacks as possible in a single product. This means that multiple antennas are packed in close proximity of each other, therefore increasing the design complexity and challenge. Fig 4 shows the evolution from a single band array to today’s state of the art multiband wideband base station antenna arrays, where antenna elements will have to co-exist with each other but are sufficiently independent with its radiated properties. Some examples of can be referenced here [4-7]

antenna elements thus inhibits the ability to provide independent tilt allowing antenna sharing for different technologies/ operator in the same bands. i.e. All frequencies having the same tilt 3.3 The solution In order to allow different access technologies (or operators) in the same/ similar band to have independent controlled down tilted antenna beams from the same array, one of the challenges is how to avoid using combining and phase processing devices at the antenna elements. A solution to this problem is through the innovation realised using a completely passive corporate feed by Quintel[11-12]. The new configuration for the base station setup is shown in Fig 5. It consists of a power divider, a phase shifter, a diplexer, and Quintel’s corporate feed for each polarisation. Each service/ operator can have independent control over its own beam tilt by controlling its dedicated phase shifter device.

Figure 4 Evolution of base station array antenna One important parameter driving the success of multiband base station antenna is the bandwidth achieved from the elements. For example, a bandwidth of 24.7% is required to cover GSM1800 and WCDMA2100, in comparison to 8.5% for GSM1800 only. Various solutions employing wideband dipoles and patches have been realised [410]. The elements of the array are typically fed via a power dividing network also known as Corporate Feed shown in Fig 3. The phase front of the antennas is then controlled via the phase shifter connected to each of the elements. This conventional method of generating phase slope to the

Figure 5 Quintel’s solution to modern base station antenna with simplified electrical tilt function

The block diagram for the corporate feed design is shown in Fig 6. The RF signals from each access technology/ operator are first translated into two equal amplitude but phase differential signals via a phase shifter. The signals from different access technologies are then recombined via diplexers and feed into the differential inputs of the Quintel corporate feed. The vector arithmetic, similar to that of the Butler Matrix[13], translate each technology’s/ operator’s signal into a phase slope across the elements resulting to antenna beam tilt. This innovation therefore allows multiple access technologies or even multiple operators (operator sharing application) to have independently controlled beams, thus opens up the prospect of multiple access technologies, or operators sharing the same antenna array. Another application of this innovation is the ability to provide Sectorisation in the Vertical Plane (SVP). This allows one operator, and one access technology to radiate two independently tilted beams, therefore achieving cell splitting within the same sector. This opens up a new dimension of cellular capacity enhancement [14].

3. Advances in Mobile Handset 3.1 Evolution to multiband handset antennas With growing requirements for connectivity in a highly mobile environment, some proposed that future mobile terminals will incorporate more than 20 separate antennas [15]. Designers are challenged to realise multi/wide band antennas within a small and premium space of the handset. Perhaps the ultimate goal is a ubiquitous antenna that could tune to any frequency of operation. Since the late 1990’s where manufacturers begin to drive design towards multiband

internal antenna solutions, numerous solutions have been reported [16-17]. We will review the challenges of handset antenna designs, followed by advances to date.

Figure 6 Block diagram of the Quintel corporate feed technology [10] 3.1 The Antenna challenges The handset antenna poses various fundamental challenges. Firstly, handsets are very small, and often a fraction of the resonance length. Radiation efficiency and bandwidth, being a function of volume occupied, would be poor with small antennas. This is especially prominent in lower frequencies such as GSM850/900 ranges. Secondly, the space in the handset is a premium. Most antennas consist of a ‘dead’ space which is mandatory for optimal performance. Failure to comply with this keep out region would mean that power is shorted (or at least coupled) out of the antenna, resulting to poor efficiency. Thirdly, being a portable device, the antenna is often shielded and detuned by the user’s

head and hand. This means that the antenna will have to have large enough bandwidth to overcome this de-tuning problem. Fourthly, regulatory authorities require that the radiation from the device complies with Specific Absorption Rate (SAR) and Hearing Aid Compatibility (HAC) directives. This poses additional challenge to the antenna design. Finally, with globalisation or more spectrums bands released, antennas will not only have to operate with current wireless systems but also cover the new and existing spectrum bands, and within the same antenna volume of present designs. 3.2 The handset cellular antenna solution Resonance characteristics of a handset antenna are critically coupled to the dimension of the chassis [18]. Especially at frequencies such as GSM850/900, the antenna is not longer the key radiating device but a driver to excite the chassis. This implies that the chassis is the ‘antenna’ hence careful design will improve bandwidth and efficiency. This eluded researches to engage in chassis driven antenna solutions shown in Fig 7a. Slots were later introduced to generate additional bandwidth as shown in Fig 7b, and evolution to a very low profile wideband antenna element using a typical volume of 5x7x45mm[19]. The typical input return loss of the antenna is shown in Fig 7c. In this case, a single antenna can replace up to 5 different antennas.

(c) Figure 7 Advance handset antenna (a) Chassis driven antenna (b) Multiband chassis driven antenna (c) Performance of integrated chassis driven antenna Tuneable antennas are also techniques used to extend the resonance range of the limited bandwidth of the antenna. Most design employs variable capacitance technique to alter the impedance match. Fig8a shows an example where the coupling slot characteristics are tuned to shift the resonance of the antenna. Such design also has an added advantage of being able to retune its resonance frequency when loaded by the users’ head and hand, therefore ensuring good radiation efficiency in talk or handheld data sessions. More information on the measured improvements to the radiation efficiency can be found in [20].

(a)

(a)

(b)

Figure 9 Parasitically coupled circular disc monopole antenna (a) Implementation (b) Simulated results

(b) Figure 8 Tunable antenna technology (a) Antenna concept (b) Measured efficiency with different tuning voltages for both low and high bands.

3.4 Cognitive radio antenna handset Cognitive radio will be the next major advancement in radio technology in the coming decade. This will make use of white space spectrum, and poorly or inefficiently used spectrum. Cognitive radio antennas are likely to be required to work in two operation modes. •

3.3 The non cellular antenna solution Whilst significant efforts are focused on integrating cellular bands antenna into a single radiator, there are increasing demands to integrate non cellular services antennas on the handset. Most handsets shipped today are Bluetooth enabled with WiFi (802.11) capability on higher end models. UWB systems are slowly gaining importance in this personal area network communication. This implies that these non-cellular antennas will have to operate from 2.4GHz to 10.3GHz. In [21], the authors have proposed a reduced size parasitically coupled circular disc monopole. The antenna is placed on the top corner of a typical handset shown in Fig 9a and the simulated response is shown in Fig 9b.

(a)

(b)

•

‘Listening’ mode, where the radio monitors the airspace for available spectrums/channel. The antenna is configured to operate in a very wide frequency band. ‘Application’ mode, where the service requested by application drives the requirements of the channelized bandwidth.

Recently researchers have begun considerations to the design of reconfigurable multi-standard handset antennas [22]. While these are single band devices, research in [23] have advanced into dual independently tuneable antennas. Fig 10 shows a simple 2 element chassis driven antenna system which occupies a very small volumetric space of 40x5x7mm. The antenna is able to dynamically adjusts its operating frequency to a wide range from 400MHz to >3GHz, and capable of both coarse and fine spectrum sensing over this broad bandwidth. This is accomplished via a tuning capacitor over a selection of inductors, and composite excitation or grounding of the elements. The input Sparameters are shown in Fig 10b-d for the different. This single antenna is now able to cover not only existing radio bands but also spectrums that are yet defined.

(b)Wideband listening mode (c) Narrowband tuning mode between 0.421GHz (d) Narrowband tuning mode between 0.7-2.8GHz

4. Conclusion (a)

(b)

(c)

While Moore’s law projects the growth of silicon processing performance, and to a certain extend data traffic growth. The rate of the number of base station antennas also increases with increasing traffic demands, but in a much more linear fashion as opposed to Moore’s exponential “demand”. This non-exponential growth has been quelled through progressive antenna innovations but which must still obey Maxwell’s laws. The term ‘number of antennas’ is used conditionally. A single antenna can be accepted so long as all the antennas can be arranged within the same space of the physical limitation of the lowest resonant antenna. Some of the latest advances in antenna technology have been described. Base station antenna sharing for multiple access technologies or even operators is now an attractive and viable proposition, including capacity enhancement with Sectorization in the Vertical Plane (SVP) from Quintel. Mobile handset will see smart cognitive systems switching to different operator bands with only a single antenna The answer to how many antennas do we need, is perhaps more Innovation.

5. References:

(d) Figure 10 Dual independently tuneable Cognitive radio antenna (a) Antenna design

1) Joyce, R.M. Barker, D.E. McCarthy, M.A. Feeney, M.T. “A study into the use of polarisation diversity in a dual band900/1800 MHz GSM network in urban and suburban environments” IEE NCAP 99, pp 316-319.

2) FCC, “Spectrum Policy Task Force Report,” ET Docket, No. 02-135, pp. 3553, released November 7, 2002. 3) http://www.sdrforum.org/ 4) Song, C.T.P., Mak, A., Wong, B, George, D, Murch, R.D., “Compact Low Cost Dual Polarized Adaptive Planar Phased Array for WLAN” IEEE AP-S Trans. vol.53, no.8, pp. 2406-2416, 2005 5) Plet, J.; Colombel, F.; “Multiband Telecommunication Antenna”, US Patent 6646611B2, 2003 6) Lindmark, B.; Lundgren, S.; Sanford, J.R.; Beckman, C. “Dual-polarized array for signal-processing applications in wireless communications” ; IEEE AP-S Trans, vol.46, no.6, pp.:758 – 763, 1998 7) Mostafa, J., “Multiband Dual-Polarised Antenna System” European Patent, EP20040805601, 2006 8) Granholm, J., Woelders, K., “Dual Polarisation Stacked Microstrip Patch Antenna Array with Very Low CrossPolarisation”, IEEE AP-S Trans vol.49, no.10, pp.1393-1402, October 2001. 9) Suh, S-Y., Waltho, A.E., Nair, V.K., Stutzman, W.L., Davis, W.A.’ “Evolution of broadband antennas from monopole disc to dual-polarized antenna”, IEEE AP-S Symposium, July 2006, pp.:1631 – 1634 10) Song, C.T.P., Mak, A., Wong, B, George, D, Murch, R.D “Multiband branch radiator antenna element”, US Patent US2005/6975278B2) 11) Haskell, P.E., “Phased array antenna system with adjustable electrical tilt”, US Patent 2006/0208944A1 12) Thomas, L.D.; Haskell, P.E. “Phase array antenna system with controllable electrical tilt”, US Patent 2007/0080886A1 13) J. Butler, R. Lowe, “Beam Forming Matrix Simplifies Design of Electrically

Scanned Antennas,” Electronic Design, April 1961. 14) http://www.quintelsolutions.com/lp/svp wp.php 15) Vainikainen, P.; Holopainen, J.; Icheln, C.; Kivekas, O.; Kyro, M.; Mustonen, M.; Ranvier, S.; Valkonen, R.; Villanen, J, “More than 20 antenna elements in future mobile phones, threat or opportunity?”, EuCAP 2009, 23-27 March 2009, pp. 2940-2943. 16) Hall, P.S., Lee, E., Song, C.T.P., "Planar inverted F antenna”, book chapter in “Printed Antennas for wireless communications”, Wiley 2007. 17) Song, C.T.P., Hall, P.S., GhafouriShiraz, H., Wake, D., “Multi-band planar inverted F antenna ”, Electronics Letters Vol 36, No.2, pp112-114, 2000 18) Vainikainen, P., Ollikainen, J., Kivekas, O., and Kelander, I., “Resonator-Based Analysis of the Combination of Mobile Handset Antenna and Chassis" IEEE AP-S Trans , vol. 50, no. 10, pp. 14331444, Oct 2002. 19) Carlo, D., Giogi, B-B., Song, C.T.P., “Wireless communication device with a multi-band antenna system”, US Patent 20090153423 20) Song, C.T.P.; Pfuhl, N.; Yuan Tao; Sager, M.; “A novel next generation self adaptive smart handset antenna”, IEEE AP-S Symposium, Jun 2009, pp.1-4 21) Song, C.T.P., Ooi, S.L., Koh, B.P., "Planar dual mode antenna for 2.4GHz and UWB", IEEE AP-S Symposium, Boston, 2008. 22) D. Manteuffel, M. Arnold, “Considerations for reconfigurable multi-standard antennas for mobile terminals”, IWAT 2008, pp.231-234. 23) Song, C.T.P., "Reconfigurable antenna", UK Patent application GB0918477.1, 2009.

Peter Song received the Ph.D. degree in Novel Antennas for Future Wireless Communication Systems from The University of Birmingham, U.K., in 2001. He was with the Applied Science and Technology Research Institute, Hong Kong, as the Manager of RF Antenna Systems overseeing development of smart antenna systems. In 2004, he joined Sendo Limited, Birmingham, U.K., and was later acquired by Motorola. He served as Advanced Antenna Specialist (Principle Staff Engineer) where he leads the mobile handset antenna research and development. Since 2009, Peter is Principle RF Antenna engineer with Quintel Technology, a leading innovator in the design, development and delivery of network-efficient cellular base station antenna sharing solutions for wireless operators worldwide. He is also holding an Honorary Research Fellowship at the University of Birmingham where his research is focused on cognitive and tuneable mobile handset antenna system. His research interest is on the development of novel antenna systems from base station to mobile handset, including multiband multipolarized wideband antennas, smart reconfigurable antennas systems, corporate feeds, antenna packaging, and various planar microwave circuits. Peter has authored over 35 technical papers and a book chapter on antenna design. He also held 3 US patents and 3 patent applications pending. He has received the Motorola’s President’s award in 2007, and Gordon Tucker Prize in 1997 from the University of Birmingham. He is a Chartered engineer of the IET and Senior Member of the IEEE