1 comments on starent contribution (x31-20051205-030) january 16, 2006 jun wang, qualcomm inc...

18
1 Comments on Starent Contribution (X31- 20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2 and its Organization Partners to incorporate text or other copyrightable material contained in the contribution and any modifications thereof in the creation of 3GPP2 publications; to copyright and sell in Organizational Partner’s name any Organizational Partner’s standards publication even though it may include portions of the contribution; and at the Organization Partner’s sole discretion to permit others to reproduce in whole or in part such contributions or the resulting Organizational Partner’s standards publication. Contributors are also willing to grant licenses under such contributor copyrights to third parties on reasonable, non-discriminatory terms and conditions for purpose of practicing an Organizational Partner’s standard which incorporates this contribution. This document has been prepared by the contributors to assist the development of specifications by 3GPP2. It is proposed to the Committee as a basis for discussion and is not to be construed as a binding proposal on the contributors. The contributors specifically reserves the right to amend or modify the material contained herein and nothing herein shall be construed as conferring or offering licenses or rights with respect to any intellectual property of the contributors other than provided in the copyright statement above.

Upload: cameron-hart

Post on 14-Jan-2016

215 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

1

Comments on Starent Contribution (X31-20051205-030)

January 16, 2006Jun Wang, Qualcomm Inc

• Notice: Contributors grant free, irrevocable license to 3GPP2 and its Organization Partners to incorporate text or other copyrightable material contained in the contribution and any modifications thereof in the creation of 3GPP2 publications; to copyright and sell in Organizational Partner’s name any Organizational Partner’s standards publication even though it may include portions of the contribution; and at the Organization Partner’s sole discretion to permit others to reproduce in whole or in part such contributions or the resulting Organizational Partner’s standards publication. Contributors are also willing to grant licenses under such contributor copyrights to third parties on reasonable, non-discriminatory terms and conditions for purpose of practicing an Organizational Partner’s standard which incorporates this contribution. This document has been prepared by the contributors to assist the development of specifications by 3GPP2. It is proposed to the Committee as a basis for discussion and is not to be construed as a binding proposal on the contributors. The contributors specifically reserves the right to amend or modify the material contained herein and nothing herein shall be construed as conferring or offering licenses or rights with respect to any intellectual property of the contributors other than provided in the copyright statement above.

Page 2: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

2

Introduction

• This contribution addresses some issues brought up by Starent contribution of (X31-20051205-030)

• This contribution only addresses the Starent’s comments on Qualcomm contribution.

• Qualcomm Comments on Starent’s Comments are marked in Red

Page 3: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

3

CDMA Network Evolution – Goals, Guidelines & Comments

Starent Networks grants a free, irrevocable license to 3GPP2 and its Organizational Partners to incorporate text or other copyrightable material contained in the contribution and any modifications thereof in the creation of 3GPP2 publications; to copyright and sell in Organizational Partner's name any Organizational Partner's standards publication even though it may include all or portions of this contribution; and at the Organizational Partner's sole discretion to permit others to reproduce in whole or in part such contribution or the resulting Organizational Partner's standards publication. Starent Networks is also willing to grant licenses under such contributor copyrights to third parties on reasonable, non-discriminatory terms and conditions for purpose of practicing an Organizational Partner’s standard which incorporates this contribution.

This document has been prepared by Starent Networks to assist the development of specifications by 3GPP2. It is proposed to the Committee as a basis for discussion and is not to be construed as a binding proposal on The 3GPP2 Secretariat. Starent Networks specifically reserves the right to amend or modify the material contained herein and to any intellectual property of Starent Networks other than provided in the copyright statement above.

TSG-X

Source: Kuntal Chowdhury: [email protected] Harper: [email protected]

Page 4: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

4

Network Evolution - Goals

• Starent suggests that before any cdma2000 network re-architecture task is undertaken by 3GPP2, a set of clear problem statements with the following criteria must be defined:– Illustrate the problem(s) clearly, preferably with real life data from

the field– Provide explanation of why the problem(s) cannot be solved

within the current network architecture– Demonstrate clear benefits of the proposed new components or

proposed changes in the current architecture

[QCOM]: It has a clear problem statement in QCOM contribution.

Page 5: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

5

Guidelines • If a new architecture is absolutely necessary, the

following guidelines must be followed:– The new architecture MUST not keep interfaces (e.g.

BS to AGW/CAP interface) closed to stifle competition among vendors

[QCOM]: We are ok with it. Actually theA17 interface specified in TSG-A can be used for the interface between CAP and the BS.

– Avoid introduction of additional network elements to serve the same functionality as in the current architecture

[QCOM]: We are ok to say: “Avoid introduction of additional 3GPP2 specific network elements to serve the same functionality as in the current architecture.”

Page 6: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

6

Comments on the Current Proposals

• At this time there are two contributions that propose architectural changes in the cdma2000 network.

• The subsequent slides include comments on these proposals

• Starent’s Comments are marked in Blue

Page 7: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

7

X31-20050926-035R1: Cisco• Problem Statement:

– Network complexity due to different levels of mobility

• Is the intent here to say: complexity due to multiple layers of mobility? In today’s architecture, mobility in the RAN does not necessarily involve L3 mobility. The PDSN insulates the core IP network from inter-RAN mobility. Depending on PDSN’s footprint, L3 mobility events may not even occur

– Scalability issue due to tight coupling between bearer data and radio link handoff

• We don’t understand this problem statement. Please clarify

Page 8: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

8

X31-20050926-035R1: Cisco• Components of the proposal :

– PPP-free operation…IP to the edge (RNC/BS): • Why a new architecture is required to achieve this?• Optimized session setup and transport is feasible with

PPP based signaling as demonstrated in AltPPP scheme

– Localized mobility and fast handoff solutions: • The current architecture is based on localized mobility

management by the PDSN and the PCF (A11/A10)• Fast handoff is also standardized as part of current

architecture• Optimizations can be made to fast handoff procedures

w/o any major change to the CDMA architecture

Page 9: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

9

X31-20050926-035R1: Cisco• Components of the proposal :

– End-2-end QoS (of course using RSVP): • Why a new architecture is required to achieve this?

• Moreover, a work item already exists on e2e QoS

– Open RAN architecture: de-couple BS and RRM: • Current Abis interface spec can be revised for HRPD

• This is a generic and a very necessary requirement regardless of the new architecture considerations

– Allow native IP host stacks as in wired networks*: • Please clarify the context: IPv6?

Page 10: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

10

X31-20050926-035R1: Cisco• Components of the proposal:

– RLP termination at the AGW: i.e. AGW directly connecting to the BS:

• RLP termination at the AGW is a major step

• Please explain the rationale

– New element: LMA for micro mobility: • The function of this entity is similar to the PDSN in

current architecture

• PDSN insulates the core IP network (HA) from RAN handoffs…something that LMA seems to achieve

Page 11: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

11

X31-20050926-029 R2: Qualcomm

• Problem Statement:– Network complexity due to too many interfaces, nodes etc.

• In the current architecture there are 7 nodes (AT, BS, RNC, PCF, PDSN, HA, AAA) and 7 interfaces (Abis, A8/A9, A10/A11, Pi, A12, A13, P-P)

[QCOM]: 5 of 7 nodes are 3GPP2 Specific. In addition, DHCP is also needed for IS-835-D deployment. Therefore, there are totally 8 functional nodes.

• In Qualcomm’s proposed architecture there are 7 nodes (AT, BS, CAP, LMHA, HA, AAA, DHCP server) and the interfaces are not numbered, but likely add up to something similar as in the current architecture (depending on the handoff and authentication schemes)

[QCOM]: Only of 3 of 7 nodes are 3GPP2 Specific.• Qualcomm’s proposal introduces a new node: LMHA• A simpler way to reduce number of nodes will be to collapse PCF with PDSN [QCOM]: We are ok to collapse RNC/PCF with PDSN as long as we have an efficient

way for mobility management for both Simple IP and Mobile IP when the MS is move from S-RNC/PCF/PDSN to T-RNC/PCF/PDSN.

Page 12: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

12

X31-20050926-029 R2: Qualcomm

• Problem Statement:– Latency (signaling and bearer)

• We definitely request Qualcomm to substantiate this problem with the current architecture. A real life data/call log with signaling latency for each of the segments: radio connection, PPP connection, MIP connection (may be TCP connection) will be helpful in understanding this problem

[QCOM]: For signaling, we know PPP introduces a delay for initial setup. For bearer, once the inter-PDSN handoff occurs, the P-P interface may be ok for current network but it does not work for inter-technology handoff.

– Inefficiency due to unnecessary bw utilization for HDLC-like framing• HDLC and PPP header less operation is possible with current network

architecture• This type of operation is already being standardized with current

architecture• No drastic change in the current architecture is required; this can be

supported in the current architecture with some changes[QCOM]: Yes if we have PFO architecture, this problem is resolved.

Page 13: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

13

X31-20050926-029 R2: Qualcomm

• Problem Statement:– Excess RAN to PDSN coupling

• Qualcomm needs to clarify this point. What does _Excess_ mean vis-à-vis what is defined in HRPD/1x IOS today?

[QCOM]: This includes several perspectives: Accounting; Flow Control; Header Compression State Synchronization due to Packet dropping at the RAN.

– Only a single channel for ROHC due to PPP• This is addressed with SO67

• Hence, why a new architecture is required to solve this issue?

[QCOM]: yes this is solved by SO67. But other issues are not fully resolved.

Page 14: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

14

X31-20050926-029 R2: Qualcomm• Components of the proposal :

– Collapse AGW, PCF, RRM into a new entity called CAP• The CAP - BTS interface MUST be standardized[QCOM]: We are ok with it if the group agree with it. Anyhow it is

not TSG-X decision.• This type of collapsing reduces network deployment flexibility:

operators need to deploy new aggregation points in the network e.g. LMHA

[QCOM]: LMHA is just a standard off-shelve HA. • If AGW and RNC are collapsed into a single node, a standard way

to manage the combined node will become necessary[QCOM]: I am not clear on what is the problem? Please clarify the

“manage” (e;g. SNMP MIB?).

Page 15: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

15

Issues w/ a Distributed Architecture• In theory, distributed processing is better, but in practice it is

not• A distributed architecture is difficult to manage and monitor• Reliability of low-cost distributed nodes cannot be matched

with super-hardened high-density nodes• Network visibility is difficult or not possible in the distributed

model[QCOM]: For this slide and the following two slides, issues with

distribution architecture are discussed. We think maybe in certain cases, the distribution architecture may be better; while in other cases, the centralized architecture is better. The point is the new architecture should allow both distributed and centralized architectures. We can let operators decide which one they want to deploy.

Page 16: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

16

Issues w/ a Distributed Architecture• Resource utilization is better/simpler in a

high-density node:– Address pool management– Better statistical gain -> less hardware

• Locating/debugging a subscriber’s session becomes extremely difficult

• Network upgrade e.g. s/w upgrade to all the distributed nodes is an operational nightmare

Page 17: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

17

Issues w/ a Distributed Architecture• Lack of standard management interface to the

distributed nodes– No standard way to manage BSCs/RNCs, where as

AGW/PDSNs support standardized management interfaces

• Data/Log collection processing: Management systems have to collect, collate and synchronize info from large number of distributed nodes

• Handoff management is not as efficient in distributed nodes – Sessions will incur more L3 handoffs therefore, more

nodes required to fix the L3 handoff latency which is introduced by the distributed architecture

Page 18: 1 Comments on Starent Contribution (X31-20051205-030) January 16, 2006 Jun Wang, Qualcomm Inc Notice: Contributors grant free, irrevocable license to 3GPP2

18

Closing Remarks• The current proposals for re-architecting the

cdma2000 network seem premature and lacking clear problem descriptions

• Most of the CDMA operators migrated their IP networks from distributed model to regionally aggregated model….

• The current network model is stable and it works[QCOM] We are not sure that we can agree with all

these statements. We need to discuss them further and agree with the problems we need to solve with current architecture.