2011 was another milestone year in the evolution of mobile communications, as the first fourth-generation (4G) Long Term Evolution (LTE) networks became commercially available. As of now, there are 17 LTE networks in operation throughout North America, Europe, and Asia. According to the Global Semiconductor Alliance's (GSA's) July 2011 report, titled "Status of the LTE Ecosystem," 45 manufacturers have announced 161 LTE-enabled consumer devices. This represents 155% growth in the number of products in development over what was originally reported by the GSA in February 2011.
Such growth is due to the nearly insatiable demand among consumers for faster mobile-broadband connections. These connections support applications like streaming high-definition (HD) video, video chat, and peer-to-peer gaming. By 2015, the average mobile-communications customer will increase data consumption by more than seven times over what they use today, according to Strategy Analytics' October 2010 report, "Global Smartphone Forecast."
However, history has shown that it can take years before a new network technology's initial launch reaches critical mass globally (over 50 million users). LTE is likely to be no different. Despite LTE's initial successes, there is still much work to be done in order to ensure a smooth global transition from third-generation (3G) to 4G network technologiesand make sure that consumers are getting the full benefits of an LTE connection. In the meantime, LTE will provide a parallel evolution path to 3G. Essentially, LTE networks will offer a high-performance, data-only standard that serves as a complement to existing 3G voice and data networks.
Imagine the global mobile-broadband network of the near futureperhaps three to five years from now. In areas where LTE was first deployed, LTE adoption continues to expand. Carriers are launching the advanced topology networks that support the features necessary to deliver strong LTE radio signals and data rates to consumers in challenging user environments. Examples include underground subway platforms and large office buildings. These advanced topology networks will continue to grow in signal strength and bandwidth. Progressively, they also will integrate features specified in later releases of LTE, such as VoIP, eMBMS, and LTE Advanced (LTE-A).
When we expand this vision of a future global broadband network into areas of the world where mobile-broadband technology is not as mature, it is likely that we will see technologies like the following still in use: WCDMA, HSPA, HSPA+, DC-HSPA+, and EV-DO Rev. A or B. They will be running on networks that are years away from supporting LTE. To ensure a smooth transition between 3G and 4G network technologies, network operators will need to "grandfather in" support for the older technologies used in less-developed areas of their networks. This mix of network technologies will require both networks and the devices operating on them to support multiple wireless technologies operating on multiple RF bands.
This ability is vital to the continuing evolution of mobile broadband, as it will ensure the best possible user experience for consumers. A smartphone user will expect a seamless and simultaneous voice and data experience while moving between 2G, 3G, 4G, and wireless-local-area-networking (WLAN) networks (Fig. 1). The phone will select and switch to the best network available at any given time.
However, supporting multi-mode and multiband operation is a complex technical issue on many levels. Today's world phones support anywhere from two to three different bands of 3G and four different bands of second-generation (2G) radio frequencies. And the number of required 3G bands is scheduled to grow to five in the very near future. This number will continue to grow rapidly with the introduction of LTE and LTE roaming, as these technologies are being deployed over a large number of RF bands worldwide. For example, some countries will repurpose bands in the 700-MHz region from broadcast television to mobile-broadband Internet standards like LTE. These are completely new bands for cellular usage. Support for them will need to be added to future smartphones.
Meanwhile, other regions will be rolling out LTE on existing 3G spectrum. This means that handsets will need to support LTE and 3G on the same frequency band (depending on which country they're being used in and which operator's network they are using). In light of situations like these, it's no wonder that the number of RF bands that could potentially be used for LTE implementation worldwide is currently hovering around 40. One also must consider the need to support the RF bands used by other technologies, which are commonly implemented in cell phones (GPS, WiFi, Bluetooth, FM radio, etc.). It becomes very apparent that it is extraordinarily complicated to support multiband RF in smartphones.
Multiband RF support also is important for LTE, as the standard was designed to have scalable bandwidth. Whereas the bandwidth for WCDMA is fixed at 3.84 MHz, channel bandwidths of 1.4, 3, 5, 10, 15, and 20 MHz have been defined in the Release 8 specifications for LTE. For a 20-MHz channel, the bandwidth that each user is allocated can vary in size from 180 kHz to 18 MHz. An LTE smartphone therefore needs to support these RF bandwidths. With the LTE Advanced part of LTE Release 10, these channel configurations are extended even further. A feature called carrier aggregation is introduced, which allows the combining of two or more bandwidth allocations.
The large number of RF bands used by LTE also creates an issue with spectrum crowding. With earlier broadband wireless technologies, a smaller number of RF bands were used. As a result, they were widely separated and interference was not much of a problem. Additionally, there was sufficient guard band for filter roll-off to provide sufficient attenuation as needed. Now, there is an increased number of RF bands in use. In addition, some of the newly designed RF spectrum for LTE neighbors existing services, such as Global Positioning System (GPS) and emergency services. As a result, the bands are growing closer and closer to each other, making interference issues quite common.
Continue on Page 2
For example, consider Band 7 and Band 38. They have both been deployed in the same areas in parts of Europe because in some countries and administrations, frequency spectrum is defined by appropriate regulatory bodies for a particular use. An example is broadband cellular communication. The spectrum is then allocated to different carriers through one method or anotherthe most common being auction. Different carriers can then deploy the appropriate services in their newly acquired spectrum pursuant to the restrictions and conditions set forth by the regulators. But what happens when different operators win different parts of the spectrum and attempt to use them for similar purposes? This is exactly what is happening with LTE deployments on Band 7 and 38 in Europe.
Band 7 is a frequency-division-duplexing (FDD) LTE band that uses 2500-to-2570- MHz uplink frequencies and 2620-to-2690-MHz downlink frequencies. In contrast, Band 38 is a time-division-duplexing (TDD) LTE band, which sits immediately in the duplex gap of Band 38 at 2570 to 2620 MHz. Interference occurs where the two bands meetspecifically, the 2570-MHz and 2620-MHz boundaries. At the 2570-MHz boundary, Band 7's uplink is adjacent to Band 38's downlink.
Because Band 38 is TDD, both uplink and downlink functions are in the same channel on the same frequency. The interference between the two bands manifests at the base-station level, where the Band 7 base station is receiving an uplink signal while a neighboring (or even co-located) Band 38 base station is possibly broadcasting. At the 2620-MHz boundary, the situation is reversed: Interference occurs at the consumer device level, when two competing devices are in close proximity (i.e., across a conference-room table, sitting next to each other on the same train, etc.). The Band 7 device is trying to receive a signal at -98 dBm while another device is transmitting at maximum power on Band 38. The result is an increase in the signal noise level for the Band 7 device.
In addition, the strength of the Band 38 signal creates additional noise for the Band 7 receiver, due to nonlinearities. Ideas have been put forth to address this issue (the use of guard bands, lowered emission requirements, sharper filtering with temperature and process compensation, better linearity, etc.). Yet this issue is still being discussed by the standardization bodies. A firm solution has yet to be agreed upon.
The interference/coexistence challenge is not limited to overlapping LTE RF bands. It also occurs when LTE comes into conflict with other wireless technologies. For example, Bands 38, 40, and 41 for LTE TDD are relatively close in frequency to the industrial, scientific, and medical (ISM) band (covering 2402 to 2480 MHz). Bluetooth and IEEE 802.11x wireless-local-area-networking (WLAN) devices also currently operate in that band. It is increasingly commonplace that both Bluetooth and WLAN functions are present in today's handsets along with the cellular radio. Therefore, there are interference conditions not only between adjacent devices as described previously, but within the device itself. For example, a user may be on a phone call using a Bluetooth headset. Or, he or she might be operating the mobile hotspot feature of the smartphone, whereby the WLAN provides a data connection to the Internet and the phone acts as a WLAN access point.
Because Bands 40 and 41 are closest to the ISM band, they are most susceptible to interference with and from ISM devices. Band 40 extends from 2300 to 2400 MHz, so it is adjacent to the lower end of the ISM band. For its part, Band 41 extends from 2496 to 2690 MHz and is adjacent to the upper end of the ISM band. The interference issue between the ISM and Band 38 is not as severe, as the bands are far enough apart that interference can be alleviated by filtering. In the past, many companies have solved the co-existence issues between Bluetooth and WiFi. Because the industry and the standardization bodies have not reached a consensus as to how to officially address these interference conditions, companies will again have to discover or invent technologies to solve the co-existence issues between LTE, WiFi, and Bluetooth.
Overcoming these multiband and multi-mode challenges will be critical for smartphone original-equipment manufacturers (OEMs). Multiband and multi-mode support will allow them to develop high-end global phones that ensure that consumers will be connectedno matter where in the world they use their phones. By designing phones using chipsets with multiband/multi-mode support, OEMs can reduce the number of device versions that they need to develop in order to support the different RF bands used throughout the global marketplace. As a result, this approach results in faster time to market while lowering nonrecurring-engineering (NRE) costs.
But today's smartphones need to be stylishand most of all, as thin as possible. Even as screens are getting larger to deliver an enhanced user experience, the volume for the core electronics within the phone is getting smaller. To attain a thin design, most of a new device's design footprint is devoted to the screen and battery.
As a result, smartphones will need to integrate all of their multiband and multi-modem support onto a single chipset (Fig. 2). Such a device needs to support multimode 2G, 3G, and 4G operation. In addition, the device must allow OEMs to implement a multimode RF front end to support advanced modems for CDMA, UMTS/HSPA, and LTE (TDD and FDD). Due to LTE's support for carrier aggregation, the transceiver must work with variable bandwidths at 1.4, 3, 5, 10, and 15 to 20 MHz. It will then have complete scalability as networks grow.
This device also should support a MIMO downlink in order to attain optimum system performance. In doing so, it will increase the overall coverage area for the mobile device and provide more capacity to the network. To be compatible with all RF bands currently deployed worldwide, such a device must support RF bands from 700 MHz to 2.6 GHz. Lastly, the device should feature an integrated GPS receiver with Global Navigation Satellite System (GNSS) and Glonass support in addition to cellular capability. Such support should be optimized for use in a cellular environment.
Implementing a multi-mode/multiband chipset is not a simple task, as it requires many key technologies (Fig. 3). On the receiver side, for example, the need to operate over broad frequency ranges with multiple interference scenarios requires the need for a high-linearity, low-noise downconverter. Reconfigurable filters are required on both the receiver and transmitter to meet the variable bandwidths of the networknot to mention low-noise broadband frequency synthesizers. Adding functionality like this is sure to increase the chipset's power consumption in addition to the overall cost of the device. To ensure that the current consumption is low, the chipset has a small footprint, and it can be available at a price point suitable for a consumer device, the key is to implement the chipset in low-cost, low-power CMOS.
Clearly, LTE is emerging as the technology of choice for future mobile-broadband networks. Yet there are many challenges the industry must address in order to achieve worldwide, ubiquitous LTE access with one devicechallenges that will take time to resolve. If carriers, device OEMs, and chipset manufacturers work together at a global level, however, LTE will provide fast but stable network connections. These will enable mobile-broadband experiences of which we have not even conceived.