Tuesday, July 19, 2005

Holiday digest: broadband considerations

The big question in broadband is: will there be room for yet another type of infrastructure, beside the incumbent technologies in both fixed and wireless:
  • Fixed: DSL (telco copper) and DOCSIS (MSO cable)
  • Wireless: GSM (Europe) and CDMA (US)

Contenders are:

  • Fixed: FTTH (fiber), BPL (powerline) and BiG (gas)
  • Wireless: WiMAX, WiFi, BoS (satellite), as well as proprietary technologies such as UMTS TDD (from IPWireless), Flash-OFDM (from Flarion Technologies) and iBurst (from ArrayComm).
The safest bet would obviously be to say: new technologies will mainly be alloted to rural and otherwise complementary deployments:

  • Its just too late. Incumbents have the infrastructure, brands, cashflows, client relations, etc.
  • There will be no regulatory support to get the new technologies off the ground, because there is sufficient competition already.
  • New technologies are as yet unproven; there are no standards yet, therefore no interoperability; also, no roaming yet; earlier attempts failed (MMDS, LMDS).
  • Existing technologies have room to grow in terms of capacity, bit rates and related costs: VDSL for telcos; DOCSIS 3.0 for cable; UMTS and HSDPA for GSM; EV-DO and EV-DV for CDMA. Sprint will even start selling EV-DO as a wireless backup link for data.
  • Despite Intel and other's support, it remains to be seen if sufficient financing is available (what ever happened to vendor financing ... ?).
  • Fiber is simply a very costly choice. However, if existing telcos (Verizon et al as opposed to Bredbandbolaget or FastWeb) choose to build out FTTH-networks, obviously the chances are much better.
  • With Nextel merging into Sprint, its iDEN technology (form Motorola) will be replaced (no definitive choice for the future yet; Flarion trials ended, IPWireless trial will start). This doesn't bode well for proprietary non-standard technologies such as UMTS TDD, Flash-OFDM and iBurst, because standards are needed for interoperability. And without that, service providers will be completely dependent on a single vendor (IPWireless, Flarion or ArrayComm, respectively).
  • Specifically, for BiG, the gas grid is not ubiquitous. And for BPL, the power grid in rural areas is not suited for broadband.

The main attractions of wireless broadband technologies are:
  • No legacy switched voice network. Providers can jump into converged offerings (voice, video and data in IP networks; fixed/mobile).
  • Cheaper than newly built fixed networks, especially FTTH.
    Much like fixed networks, wireless networks can boast ever expanding bit rates through technological breakthroughs as well.
  • While earlier standards (MMDS, LMDS) failed, WiMAX has learned a couple of lessons: it is standards-based (thus enabling interoperability between different vendors' equipment), it requires no line-of-sight (no hindrance form snow storms or flocks of birds or other disruptions) and it preferably uses licenses spectrum (to avoid interference).

My personal bet would be on cellular providers:
  • They are incumbent operators.
  • Their networks are ubiquitous.
  • They have the inherent cost advantage of wireless versus fixed.
  • HSDPS and 4G are promising.

For background reading, and also to highlight how quick things change in the broadband space, an overview of recent developments:
  • Light Reading along the way in this article implicitly assert that broadband + MPEG-4 = FTTH (MPEG-4 is the new compression standard. It requires half the bit rate of MPEG-2. One HDTV stream needs only 3-4 Mbps of bandwidth.)
  • NTT DoCoMo, on the way to 4G, have achieved 1 Gbps.
  • The new DOCSIS 3.0 standard will deliver 100 Mbps bit rates to cable operators. Interestingly, also IPTV and videoconferencing. Available early 2006.
  • The WiMAX Forum started testing a first batch of equipment, probably from Airspan, Alvarion, Aperto, Redline and Siemens. Official WiMAX equipment will be available early in 2006. All this relates to the 802.16d (or 802.16-2004) standard. Mobility will be added in 802.16e, to be ratified toward the end of 2005.
  • PanAmSat is working on a satellite-based version of WiMAX (base stations mounted on satellites). Services are to include IPTV to the entire US.
  • The next WiFi standard (802.11n), enabling bit rates of 100 Mbps, is closer than expected earlier. This way, WiFi could fend off teh WiMAX threat.
  • Japan Telecom will begin trials on a wireless network combining WiFi and Flarion's Flash-OFDM.
  • Flarion scored several deals lately (Croatia, Malaysia, Virginia, Finland).
  • Google et al invest $100m in Current Communications, which offers BPL in Ohio in conjunction with utility Cinergy.
  • Nethercomm, developer of BiG technology, will host a conference on August 24-25. It now aims to work with a telco or cable operator, but how remains unclear. Cooperation with an unaffiliated ISP (such as AOL or EarthLink) seems to make more sense.

No comments: