Channels allocated for wireless services are increasingly being crammed together on the frequency band while the price per MHz of spectrum remains quite high at auction. This is especially at issue for 4G signals, which are commonly located at frequencies adjacent to existing WiFi and Bluetooth channels and therefore suffer from mutual interference issues. As such, operators and device makers need interference mitigation solutions. One such solution arises out of an advanced filter technology known as bulk acoustic wave (BAW) that has emerged over the past few years. This article will investigate the driving forces behind the need for better filtering for 4G applications and discuss an effective hardware-based solution.


As demand for broadband services grows, the trend towards wireless is well-established. In emerging markets, such as China and India, build-out of wireline infrastructure will be too expensive, time consuming and logistically complicated. In established economies, mobility is the “killer app” that drives the growth in demand for wireless devices and applications. But wireless communications require spectrum, which is—in economic terms—a scarce resource; this has three key implications.

First, a limited supply coupled with strong demand will naturally lead to high prices. This has been evident in the billions of dollars generated at auction for the rights to these bands. For example, this year’s 4G auctions in India generated $8.2 B. The popularly quoted revenue figure for the India BWA auction is $5.5 B (USD), which doesn’t take into account the additional $2.7 B paid by state-run operators BSNL and MTNL. Without having to actually participate in the auction, these two firms were guaranteed spectrum at a price equivalent to the winning bid. This equated to an average of $6.2 M per MHz of spectrum, with that price per MHz per capita peaking over $1.15 in Mumbai and Delhi.

Figure 1 Allocated frequency bands.

Second, high demand for a scarce resource motivates a supplier to produce as much of that resource as possible. While it is not possible to “produce” more electromagnetic spectrum, it is possible to “refarm” existing spectrum. To that end, regulatory bodies worldwide are working to find more channels to put into use, with the assumption that interference or other considerations will be handled by the marketplace. The FCC, with the support of President Obama, has released The National Broadband Plan this year. Amongst other recommendations, the document calls for 500 MHz of new spectrum to be made available for broadband within 10 years, 300 MHz of which should be allocated for mobile use within five years. To the Broadband Plan’s credit, it does recognize that interference issues are lurking, but it does not offer solutions. Meanwhile, many bands allocated for 4G data services happen to be immediately adjacent to the International ISM band, the unlicensed band that runs roughly from 2.4 to 2.5 GHz and is used worldwide for Wireless LAN and Bluetooth signals. Another example in the US is the satellite radio band, which sits right in the middle of WCS spectrum with no guard-band defined by the FCC (see Figure 1). As band assignments get tighter, interference issues multiply.

Third, scarcity leads owners of a resource to use it as efficiently as possible. There are two main approaches to achieving spectral efficiency, both equally important. The first is the use of better modulation techniques, allowing an operator to pack more data into a given channel. The second is to find ways to stretch the use of purchased spectrum to its edges—to use an entire 20 MHz channel for data rather than give up 5 MHz on each side as guard-bands.

Let us take a moment to consider guard-bands. A guard-band is the slice of frequency between two channels, a sort of “no-man’s-band” where both parties gradually roll off their transmission power. In some cases, guard-bands are built into the operator’s license; the FCC might designate 10 MHz between channels that cannot be used. More commonly, the regulatory body tends to leave the definition of the guard-band to the operator. The requirements for out-of-channel transmissions will be set by, say, the FCC, and the license holder is free to use as much spectrum as he wants, providing he meets those rules. Guard-bands, naturally, represent wasted spectrum, and operators are intent on minimizing them.

Arising out of these implications is a clear picture: 4G operators are spending a great deal of money on channels that happen to lie very close on the frequency spectrum to interfering signals, but have no built-in guard-bands. The standards upon which 4G data services will be delivered are WiMAX, LTE and TD-LTE. These are extremely similar in terms of transmission pattern (all use the OFDMA modulation scheme); and, more importantly, there is a significant overlap in the frequencies that each will use. As such, for the purposes of this discussion, they can be referenced collectively as “4G”. While there are sub-2 GHz bands set aside for LTE (most notably, the 700 MHz “digital dividend” spectrum), a significant proportion of 4G services will be delivered in what the author calls the 4G bands: 2.3 to 2.4 and 2.5 to 2.7 GHz (see Figure 1).

Operators find this spectrum attractive for three main reasons. First, it has relatively good RF propagation characteristics. Second, it is largely unused, meaning there are bigger chunks of consecutive spectrum available here than elsewhere in the band. Third, we are seeing some level of global harmonization, with governmental regulators in India, China, Japan, America and Europe all offering this spectrum for 4G services. Alas, nothing is perfect and there is a key drawback, discussed earlier: nestled amongst these attractive 4G bands are nasty interferers in the form of WiFi, Bluetooth and satellite radio signals.

This issue is perhaps most apparent and acute in the increasingly popular “personal hotspot” devices. These are wireless LAN access points that take a 4G signal from the service provider and convert it to WiFi in order to share the 4G connection across multiple WiFi-enabled devices. Because they transmit and receive 4G and WiFi signals at the same time, these personal hotspots have the greatest risk for mutual interference issues.

Figure 2 Example filter performance.

In order to address this issue while minimizing wasted spectrum in the guard-bands, operators and device makers have turned to advanced filtering technologies. A filter with a steep skirt—one that rolls off quickly from pass-band to rejection band—becomes increasingly important in cases like these (see Figure 2). Surface acoustic wave (SAW) filters have traditionally served this purpose. However, as frequencies rise to the 2.3 to 2.7 GHz range, SAW performance starts to decline. BAW filters are the natural evolution of acoustic wave technology for higher frequency bands, and become extremely attractive as frequencies grow to 2.3 GHz and above.

For example, TriQuint Semiconductor’s BAW technology has been applied to a trio of filters specifically designed to mitigate the interference issues between the ISM and 4G bands. One filter in this family passes the ISM band while effectively rejecting signals above 2.5 GHz. One application for this part would be as a WiFi pass-band filter. Notch filters are used to knock down WiFi/BT signals and pass the 4G signals with low insertion loss both above and below the ISM band. In addition to excellent electrical performance, BAW filters can be extremely small with very low profile packaging. Filters are available in 1.3 × 1.7 mm packages with a profile of less than 0.5 mm. Clearly, this supports the long-term trends toward ever-smaller, full-featured mobile devices. BAW also exhibits excellent power handling capabilities, with WiFi notch filters typically withstanding +28 dBm (continuous wave). Used in various combinations depending on the end application, these BAW filters are good examples of how companies are simplifying RF connectivity with new hardware technology that resolves real-world problems like the interference and band adjacency issues 4G operators now face.

Demand for 4G will continue its path of explosive growth and it is conceivable that, moving forward, every scrap of spare spectrum will be prized as a vehicle for delivery of lucrative broadband data services. As useable spectrum bands continue to press together, innovative filtering solutions like BAW technology will continue to help operators make better use of their spectrum and realize more complicated and compelling consumer devices.