BROADWAY, the way to broadband access at 60GHz

One of the most important outcome of BroadWay project is to have demonstrated the feasibility of a dual band 5/60GHz system approach underlying the complementarity of the two bands in an integrated manner. The prototypes developed throughout the project have validated the relevance of the considered approach that can be viewed as an alternative in Europe for UWB that in a 3 years timeframe can become competitive. Moreover these prototypes are used and will be used as a powerful convincing tool both internally and externally for advocating 60GHz short range communication systems.

One of the key objective of BroadWay is to impact the standard bodies and increase the overall level of awareness for 60GHz short range communication systems. Two standardization paths have been explored for a landing zone for the BroadWay project: ETSI BRAN (Broadband Radio Access Networks) and IEEE802.15 for WPAN (Wireless Personal Area Networks). After the joint presentation of last year to ETSI BRAN by Motorola and FT (from another RNRT consortium ComIndoor) for pushing at the ETSI BRAN level for the creation of an work item on 60GHz WLAN/PAN it has been decided not to pursue this path, since it was conflicting with CISCO interests and requirements to use this band for point-to-point directive transmissions which was not compliant with BroadWay WPAN system concept. As a result a greater emphasis was put on the IEEE arena that is ruling for the moment the standardization of WLAN and WPAN solutions. Motorola has been key in the creation of a WPAN Millimeter Wave Alternative PHY Study Group 3c (SG3c) IEEE802.15: SG3c i.e. IEE802.15.3c (BroadWay european project mentioned by Siemens on the ieee web page: http://www.ieee802.org/15/pub/SG3c.html). Three presentations were made by Motorola at various IEEE802.15.3 meetings. Note that this study group should be transformed to a full task group during next year. The issue of this process will consist in a call for proposals for a new PHY layer at 60GHz followed by a down selection voting process to elect the proposal to be considered for standardization.

Concerning the layer 2: a novel Centralized Ad-hoc Network Architecture (CANA) was proposed in order to allow for the coexistence of a short range ad-hoc mode of operation under the umbrella of the traditional WLAN cell (HiperLAN/2 in particular). This new architecture includes all adhoc algorithms comprising node discovery and route as well as dual mode of operation between 5 and 60GHz comprising related DLC and CL enhancements. It has been verified that these DLC extensions allows a seamless operation between 5GHz (in a TTD/TDMA centralized mode) and 60GHz (using a centralized clustered architecture with Neighborhood discovery and adhoc routing) with reasonable overhead. As a result of these studies an entire software platform based on the Network Simulator tool integrating all the CANA modules has been implemented and is usable for detailed performance assessment. The development, evaluation and testing of the new dual-mode MAC layer, the required Neighborhood Discovery process and the interactions with the upper layer have been performed. The goal of integrating the ad-hoc concept at 60 GHz into the standard HIPERLAN/2 system by providing for the concurrent operation of the AP at a 60 GHz cluster has been achieved. In this way, it was shown that MTs may take advantage of the total capacity increase by using the shorter-range and more vulnerable 60GHz paths under the appropriate adjustment of the developed algorithms.

A new innovative SISO and MIMO Pseudo Random Postfix OFDM multicarrier modulation schemes hase proposed in order to be robust to the high Doppler present at high carrier frequencies such as 60GHz. This new modulation scheme is specifically designed to be robust towards mobility (channel variations). Two patents have been filed during the course of the project concerning this new modulator. The scope of application of PRP-OFDM is not restricted to 60GHz but applies at all frequencies and is well suited for lower bands for vehicular outdoor operations such as in IEEE802.11p or IEEE802.16a/b. BroadWay has been the platform to demonstrate and validate through prototyping this approach by benchmarking against more conventional technologies such as Cyclic Prefix based OFDM schemes. Moreover a complete baseband solution has been proposed and its parameters tuned for the use at 60GHz in full compatibility with the capabilities of the realistic WP4 analog front-end architecture.

One major outcome of the BroadWay project is the design of all advanced MMICs for building up the 5/60GHz up and down converter. First path on the custom design of the various advanced MMIC blocks has been accomplished producing: single stage amplifier, two stage amplifier, two stage amplifier with double output stage, branchline mixer, branchline mixer with LO buffer amplifier, image reject mixer, ratrace mixer, ratrace mixer with LO buffer amplifier, LNA, multiplier (x4) a number of teststructures for process and design analysis using WIN InP process libraries. Useful and relevant experience has been gained with using the mHEMT process of WIN. Moreover BroadWay has released a package of the 60GHz custom designed chips. The overall cost of the unit has to be kept as low as possible. This task combined with the fact that the unit is working at 60GHz is very difficult. In an attempt to reduce the cost and use the in-house available pick and place machine we used completely different approach than before. All the RF components were located on one side of the unit and all DC components on the other. This reduces to minimum the RF leakage to the DC components and also helps during the assembly of the units. The unit itself was divided to compartments in order to reduce the internal RF leakage. As a part of the overall system BroadWay has to develop a low phase noise at 56GHz. The only candidate for this was a PLDRO, which has to be multiplied after this to the required frequency. As will be seen in the next paragraph we succeed and developed a PLDRO, which could be integrated on the overall package. As a result Farran and TNO have new designs to be put available on their product catalogue (actually both the up and down converter are in the Farrans catalogue at present).

Concerning the investigation of new key antenna technologies: three new designs tailored to the BroadWay system have been fabricated and fully measured: - A 60GHz directional antenna design (high gain planar antennas) for access points - A 60GHz omnidirectional antenna - Two dual band 5/60GHz antenna design These antennas are all-available for product integration and have been packaged into the BroadWay prototype in the 5/60GHz up/down converter. When one looks specifically at the design of antennas at microwave frequencies, the following should be noted. Designing antennas at these frequencies is always a hazardous job, not with respect to the design itself but the manufacturing tolerances and losses. Small variations in material properties and layouts can cause at these high frequencies a relative large shift in bandwidth or even a drastically change in return loss. This can result in a complete malfunction of the antenna. Besides this point losses can be significant and feeding structures should therefore be as short as possible. These are normally the 2 bottlenecks for using planar antenna structures at these high frequencies. Taking into account all the manufacturing tolerances is a very elaborate task, which not always guarantees a successful outcome. Within Broadway it has been examined how these tolerances affect the performance of the antenna. One of the major outcomes of Broadway with respect to the RF-side is that with the right manufacturing strategy the tolerance influences can be controlled very well, the overall reproducibility is good, and that the losses are not as high as expected. These findings encourage companies like IMST and Motorola to push antenna designs into higher and higher frequency ranges, for applications such as higher generation mobile antenna systems (palmtops, laptops, cell phones) where miniaturization is always an key issue. Because of the high frequencies antennas become much smaller than at lower frequencies. This implies not only smaller terminals for a broad range of applications, as wells as indoor as outdoor, operating at high data rates, but also the possibility to included more than one antenna into a mobile terminal. This would open up the road to complete other applications like MIMO (Multiple Input Multiple Output) systems where various antennas are used to increase the data capacity of an RF-link. Employing such techniques would introduce a whole range of multimedia applications, thus opening up markets with a high economic potential. A complete other area of interest are sensors, for example those used in car industry at 77GHz; anti-collision radar, etc. The type of antennas used her are normally based on lens designs. Using the Broadway experience, it may also be possible to use standard planar designs that are very compact and can be easily integrated within the car architecture. The car aesthetics are one of the most important market driving forces, and planar or nearly planar structures in the microwave frequency range are almost invisible because of their small dimensions. Other applications can be traffic management systems that use a large amount of sensors along the high ways to check, for example, for congestions or to impose penalties of charges.

Risultati finali

Condividi questa pagina

Scarica