Introduction It is with pleasure that I provide my suggestions for spectrum reallocation of the digital dividend. Wireless is the solution of choice to connect far-flung regional and remote users to the Internet back bone (e.g. the NBN). However there are no present-day or proposed next generation wireless networks capable of simultaneously achieving scalable (spectrum reuse), long range (50 – 100 kms non-line-of-sight) and broadband communications. However wireless technology is advancing at a frantic pace. There is even the distinct possibility of a discrete jump in performance as researchers find new ways to reuse spectrum.
In anticipation one should therefore allow portions of the digital dividend spectrum to be dedicated for special use in both regional and remote areas. In the Discussion Paper this dedication of spectrum is referred to as excision from spectrum licensing of the digital dividend. Here we propose that this excision, or partial excision, also apply to regional areas.
As shown in Fig. 3 of the Discussion Paper, there are many regional areas (shaded in blue) which are in some sense geographically close to the National Broadband Network. The NBN map for NSW is shown in Figure 1. These still require a long range wireless broadband solution to make the last mile connection.
F igure 1. The blue lines show the fibre runs of the NBN in NSW (http://www.nbnco.com.au/)
Specific comments I confine my responses to questions 3, 4 and 5.
3. Should remote parts of Australia be excised from spectrum licensing of the
digital dividend band? Why? Certainly. As stated in the paper, given that spectrum allocations do not necessarily lead to the delivery of services in remote areas for economic reasons, it would be a pity to waste the spectrum when it could be exploited for innovative local purposes.
There is also the question of where to draw the boundary between regional and remote areas. Given that many of the areas classified as regional in Fig. 3 of the Discussion Paper could be connected to the NBN on the 700 MHz band, the government may wish to consider excising some of the digital dividend spectrum in regional areas for class licensed wireless access services.
4. If not, how much digital dividend spectrum should be made available in remote areas of Australia? Should different amounts of spectrum be made available for spectrum licensing in different areas, for example, metro versus rural? Following on from Q3, much could be achieved by assigning a 20-40 MHz (Wi-Fi equivalent bandwidth, for example) portion of the 126 MHz for general use in regional Australia.
5. What services would potential licensees of digital dividend spectrum expect to
deploy in remote Australia? Internet back-haul to the NBN in regional Australia would be a clear possibility. With the availability of TV towers and microwave back-haul much of regional Australia could be connected to the NBN using 700 MHz spectrum.
Figures 2 and 3 show the locations of all the TV towers in Australia.
F igure 2. TV towers with circles showing 20 km radius
The red circles show a 20 km radius (for clarity only 20% of these circles are drawn). However the TV sites are already chosen to achieve fairly complete TV coverage at 700 MHz. We can see that in many regional areas, the TV towers provide by design, a wireless access point on 700 MHz. The main technological problem will be to reach the NBN from the tower. Provided a microwave link (for example) can connect the tower to the NBN, then we have an ideal wireless access solution for regional Australia.
Figure 3 shows what can be achieved for a 50 km radius. To do this, one could use 10-20 meter towers at 50 MHz in normal Australian terrain. The band I TV spectrum can provide coverage in most areas of Australia and in many cases will not need back-haul to the NBN.