TECHNICAL
Some questions for further discussion:
suite and see http://www.computer- timeline.com/timeline/vint-cerf-and-bob- kahn/ for some background. My basic Bitco model examines functionality only and does not examine how bits are delivered. The Bitco model is invariably more complex than the basic model in Figure 3 opposite top and see Figure 4 opposite centre, for a more illustrative scenario. Here the interconnected networks (not functions) can be modelled as series of interconnected queues. My observations on the complexity of delivering bits between end points is bounded to fixed networks only 14 . In addition, and as my analysis is cursory, others have written at length in this area e.g. David Clark 15 and Larry Peterson/ Bruce Davie 16 with their Systems Approach thinking. For an alternative view, see John Day. 17 To aid further discussion, I have reproduced a diagram 18 (with permission) from an academic source and see Figure 5 opposite, bottom. Although there is a degree of complexity in this diagram, the “network of networks” model of an internet is evident.
a network of networks with end to end management (if any) based on best effort delivery within a network 19 . It is also worth mentioning that the term “Telco” is not mentioned within this diagram with “ISP” (Internet Service Provider) as a more appropriate term, perhaps? A Note on QoS There is an ongoing industry debate on specifying QoS for packet networks which suggests that delivering QoS across multiple interconnected packet networks is a challenge – see latest IETF work in this area. 20 In the circuit switching world, QoS is well defined and documented - the ITU approach 21 is commended. My opinion on introducing QoS within the Bitco model is an act of folly – are we trying to emulate a circuit switch approach? Observations Isenberg’s paper describes a network which could be described as the Internet and note at the time of writing his paper i.e. May 97 much of the experimental
n If the Internet has replaced the legacy telco model how is the Internet governed? n Increasingly, Government Regulators are introducing concepts such as the European Commission and its “Fair Share” initiative. Are such interventions breaking the original Internet model? n Isenberg suggests that the Internet dis-intermediates the telephone network. If so, why do we continue to describe large (legacy) internet service providers (ISP) as Telcos?
n Is the Internet now the “Stupid Network?”
Summary This paper was authored to test my assumptions/knowledge and in particular the use of the term Telco (Telecommunications Company). If we can agree that the ISP/Telco are synonymous, I can live with it but prefer ISP.
research on the protocols and infrastructure had been tested.
www.firstmilenetworks.co.uk
An observation and sometimes lost within “the chatter” in that the Internet is
REFERENCES
13 See https://www.internetsociety.org/internet/history-internet/brief- history-internet/ for an excellent summary. 14 There are many “commentators” who understand mobile networks far better than I. 15 David D. Clark: Designing an Internet, 2018, MIT Press. An excellent historical perspective. 16 Larry Peterson and Bruce Davie: Computer Networks: A Systems Approach, 2024, https://book.systemsapproach.org/ also available at https://7e.systemsapproach.org/ which is an online development version of their book – highly recommended. 17 John Day: Patterns in Network Architecture: A Return to Fundamentals, 2008, Pearson. Day’s work is also highly recommended and I must declare an interest here in that I know John Day quite well. Day’s approach is contrarian to say the least. 18 Shedding Light on the Internet: Stakeholders and Network Neutrality: Antonopoulos, et al. IEEE Communications Magazine, July 2017 19 Many academic papers have been written on the end to end delivery model and see M Handley, “Why the Internet Only Just Works”: BT Technology Journal Vol 24 No 3. July 2006 is a good starting point. 20 https://datatracker.ietf.org/doc/draft-ietf-rtgwg-qos-model/ 21 See - https://www.itu.int/dms_pub/itu-t/opb/tut/T-TUT-QOS-2023-2- PDF-E.pdf
1 The Changingman, Paul Weller, 1995 2 Isenberg’s paper was available at https://isen.com/stupid.html but is no longer available. My paper (which includes Isenberg’s original paper) is available at http://www.firstmilenetworks.co.uk/Public/vol4p144-51.pdf 3 Cf with the Energis model in the UK 4 A contentious title, perhaps? 5 To quote a past lecturer “Analyze, build a model and subsequently test (many times).” 6 Increasing Scope for Circuit Switching in the Optical Internet: ICTON 2009 7 Whilst at NatWest Bank we used a network model created by BBN for our packet network. Unfortunately, this was a long time ago and for the life of me, I cannot remember if it helped in our capacity planning process 8 Attributed to John Day 9 Data Communications at the National Physical Laboratory (NPL) -1965 to 1975” by Martin Campbell-Kelly 10 Is Telco now an agreed term for an ISP?. 11 Compare with the legacy term “Telco.” 12 For the sake of simplicity, the model illustrates bit transfer in one direction only.
Volume 48 No.23 SEPTEMBER 2026
99
Made with FlippingBook - Online magazine maker