To: Reinhard Scholl From: Patrik Fältström Cc: IAB , ITU-T Liaison coordination Subject: ITU-T Q 17/13 – Packet forwarding and deep packet inspection for multiple services in packet-based networks and NGN environment Dear ITU-T, The IAB noted that the motivation of "Question 17/13 – Packet forwarding and deep packet inspection for multiple services in packet-based networks and NGN environment" states that DPI (deep packet inspection) is essential for multi-service delivery in packet-based networks. We would like to inform the ITU-T that this is not correct. Specifically, deep packet inspection is only one of many mechanisms that one can use to identify what traffic is flowing in a network. Moreover, deep packet inspection may have negative impact on the packet forwarding performance and serious implications for privacy. Also, this mechanism is not effective in some cases. We would like to draw your attention to solutions that do not involve DPI and may offer a superior functionality. See, for example, work in the IETF LEDBAT and CONEX working groups. It might be useful to also consider different approaches, like statistical protocol identification (http://spid.sourceforge.net/sncnw09-hjelmvik_john-CR.pdf), and some deployed solutions for multi-service delivery in packet-based networks (http://tools.ietf.org/html/draft-livingood-woundy-congestion-mgmt-09). We, therefore, urge the ITU-T to specify the charter for the question differently. The current charter is already focused on the solution, namely DPI, rather than on identifying a problem and allowing other, more superior solutions to be considered. Patrik Fältström Liaison from IETF to ITU-T