LTE Policy and Charging Control (PCC)

August 22, 2014 | By Netmanias (tech@netmanias.com)

Online viewer:

HTML

Download
PDF File

Comments (27)

Page 2 of 6

Table of Contents

1. Introduction
2. PCC Rules

3. PCC Procedures

4. How Policy Control Affects the Way IP Packet Flows are Handed

5. EPS Entity Information

6. Closing

2. PCC Rules

PCRF determines a PCC rule for each SDF, and forwards them to PCEF (P-GW) over the Gx interface. P-GW then has the forwarded PCC rules enforced for each SDF. After enforcing the PCC rules, when IP packets arrive, it detects SDFs that each packet belongs to, and applies a PCC rule to each packet according to their SDF. 3GPP TS 29.212 [1] defines PCC rules as follows:

Purpose: The purpose of the PCC rule is to detect a packet belonging to a SDF, identify the service the SDF contributes to, provide applicable charging parameters for the SDF, and provide policy control for the SDF.
Apply to: Different SDFs. PCEF (P-GW) classifies IP packets by matching them against the packet filters (SDF templates), and the matching PCC rules are applied to the packets.
Type: Dynamic PCC rules or pre-defined PCC rules. Dynamic PCC rules are dynamically provisioned by PCRF to P-GW when an EPS session is established or modified. Pre-defined ones are preconfigured in P-GW, and thus can be activated or deactivated by PCRF.
Elements: A PCC rule consists of a policy rule name, service ID, SDF template, gate status, QoS parameters, charging parameters etc.,² and varies depending on the operator’s policy.

Below, we will see how PCC rules in different types (pre-defined and dynamic) are delivered over the Gx interface, and are enforced in P-GW. Section 2.1 describes the delivery of pre-defined PCC rules and Section 2.2 explains that of dynamic PCC rules.

2.1 Pre-defined PCC Rule
Some operators apply a policy that limits the bit rate of traffic not served by themselves, for example P2P. So, we will explain pre-defined PCC rules by using P2P as an example. Figure 1 illustrates how a pre-defined PCC rule is configured in P-GW.

Figure 1. Example of Pre-defined PCC Rule: P2P Traffic

In the figure, “P2P”, a pre-defined rule, is pre-configured and deactivated in P-GW. Here, because this rule is pre-defined, once PCRF decides on it, it just needs to give the P-GW the name of the policy rule (i.e. P2P), instead of sending all the elements of the rule. Upon receiving “P2P” rule name, the P-GW activates the “P2P” rule that has been deactivated, and enforces it.

2.2 Dynamic PCC Rule
Dynamic PCC rules are used whenever there is an event. Figure 2 shows how a dynamic PCC rule is configured in P-GW when a user uses Internet and voice services. The figure assumes the followings, and shows all the dynamic PCC rules to be covered in this document.

Operator policy – Internet services are accessible to the extent allowed by the network resources, and the data rates of voice services are guaranteed.
APN – Internet and voice services are accessed through different APNs, e.g. Internet and IMS, respectively.
PCC rule name – Internet service: “Internet”, voice service: “Voice-C” (SIP signaling)/”Voice-U” (voice packet)

Figure 2. Example of Dynamic PCC Rule: Internet/Voice Service

“Internet” rule applies to all packets that access the Internet. It supports unlimited MBR (UL/DL), allowing maximum available bit rate for the Internet access. This rule applies to the default bearer accessing the Internet. APN-AMBR (UL/DL) is set unlimited, and data travels at maximum rate if there is no other user being served. But, the more users are served, the lower rate is supported.

“Voice-C” rule applies to SIP signaling packets. It applies to the default bearer that accesses the IMS network, supporting APN-AMBR (UL/DL) of 100 Kbps.

“Voice-U” rule applies to media packets, which are user voice packets, and supports GBR. During an active voice session, the network resources are allocated to ensure that packets are delivered at UL/DL 88 Kbps all the time. Also, a dedicated bearer is established, allowing packet delivery at GBR (UL/DL) 88 Kbps.

Every time an EPS session is established or modified, PCRF dynamically determines a PCC rule including the rule name (e.g. “Internet”, “Voice-C” or “Voice-U”) and the policy parameters specific to the particular rule depending on the service type selected. Next, it provisions them to P-GW through the Gx interface. Then the P-GW enforces the received policy parameters to IP packets after mapping them to bearer QoS parameters.

Once a user accesses the Internet and IMS network through initial attach procedure, the user’s default bearers are kept connected unless he is detached from the LTE network. So, basically, “Internet” and “Voice-C” rules are kept active in the P-GW. On the other hand, a dedicated bearer on IMS APN is established upon detection of a voice call, and terminated as soon as the call ends.

Page 2 of 6

Emmanuel Pattinson 2014-08-31 18:56:31

Thanks Netmanias, truly appreciate your efforts in preparing this document.

Any update on Online and Offline charging. (OCS and CGF)???

Thanks mate!!!

Raj 2015-11-28 15:24:51

Who assigns the Dedicated bearer ID ? not clear from document. Is it MME or PGW

Chris 2015-11-28 17:24:43

@Raj, MME assigns the EPS Bearer ID (EBI) for both default and dedicated bearer.

Saurabh Thukral 2014-09-23 16:20:46

refer Figure 4.

An important parameter missing in the call flow for voice call modification procedure i.e setup of dedicated bearer is LBI. Please add this parameter

Netmanias 2014-10-10 18:40:32

Hi Saurabh,

Thank you for the comment. We have added EPS bearer related IDs to Figure 3 and 4 (Default Bearer ID in Figure 3, and Dedicated Bearer ID and LBI in Figure 4), and the related STEPs have been modified (STEP 2) in Section 3.1, and STEPs 6) ~ 7) in Section 3.2).

We omitted EPS Bearer IDs and LBI to simplify the figures because the IEs are mentioned often in other LTE TRs. But, as you pointed out, LBI is important parameter when a dedicated bearer is created, so it seems better to add them. Thank you for your careful comment.

Abhishek Jadhav 2014-09-27 13:45:33

It is clear realization of technology

Thank you very much

Lavanya Sabhapathi 2014-10-09 13:26:30

Thank you so much for sharing wonderful document which made me to understand about default and dedicated bearer. If you can plz share about messages like ASR STR clearly

Harry 2014-10-24 04:01:29

Thank you so much for this excellent document.

What is unclear to me is, in case of PCC rule with closed gating status, whether or not the gating-status is delivered to SGW and/or UE as part of TFT template. I'd like to understand where the blocked uplink packets are discarded. (PGW or SGW or UE?)

Thanks again.

Vicent Ferrer Guasch 2014-10-29 00:57:47

Hi,

First of all, congratulations for all these amazing LTE technical documents. They are really clear and helpful in order to understand the complexities of LTE, in addition to a nice quick reference. Thanks a lot for sharing.

I have a question about figure 6. The Bearer ID for both Default bearers is 5, is that correct or is it a typo? Is it possible to reuse the Bearer ID for different APNs?

According to TS 23.401 clause 5.2.1, "An EPS bearer identity uniquely identifies an EPS bearer for one UE accessing via E-UTRAN"

Is there any limitation on the number of Default Bearers that can be stablished, apart from the 11 Bearer IDs available?

Netmanias 2014-11-01 10:19:09

Hi Vicent,

It’s typo error. As you have mentioned, different bearer id is allocated to each bearer for an UE. We have corrected the typo in Figures 6 and 7 and updated this post and pdf file. Thank you for letting us know that.

We are not sure about the maximum number of default bearers for an UE, but it may depends on the number of PDNs the UE can be connected to. Some eNBs/femto APs have been implemented to support only a limited number of PDNs.

Antti Mikola 2014-11-05 02:19:35

Hello,

Thank you for this informative document.

I have a question concerning the dedicated bearer creation. Can two bearers with the same QCI values but different ARP values exist at the same time? For example, concider the following scenario. A new PCC rule is authorized with a certain QCI and ARP values. If a bearer with the same QCI but with a different ARP value already exist, can the PCC rule be mapped into this bearer or is a new bearer with same QCI but different ARP value created?

radhi 2015-03-24 19:57:34

No, Bearer is uniquely identified by the combination of QCI and ARP. So even if ARP is different, dedicated bearer creation will be triggered by PCEF.

pri 2014-12-18 23:31:51

How does the PCRF store the PCC Rules (based on what index). I mean lets say same QOS is applied

for two Applications (mapped to SDF1 and SDF2, both mapped to dedicated bearer id 1.

So lets say the App related to SDF#1 is stopped. Then PCRF needs to delete PCC Rule related to SDF#1.

But the rule is still valid for SDF#2. Any pointers on how this is implemented in PCRF.

It will be great if you can give some practical example (with just 1 or two rules in PCRF) .

Also what will be sent in the Diameter messages esp in Gx interface... Also may be bit about Rx interface ?

vikas singhal 2015-03-30 18:14:52

Hello,

Thanks for such detailed documentions, I guess there is a typo in Figure 6. For PF1 of Deafult bearer (DL), SDF id is written as SDF2 which should be changed to SDF1.

ashish12784 2015-08-08 19:15:13

I have some doubts related to section 3.1.

1. Suppose UE have configured multiple apns (internet.apn.com, ims.com) for internet and ims services. UE is switched on, shall it create eps session with both the apns or any of default apn (internet) and eps session with ims once UE attempt any voip call?

2. Suppose there is only single apn (ims.com) is configured in UE, Is pdn connectivity request need to send from UE to MME or attach shall be sufficent?

3. As per my understanding, there is one eps session for each apn. Shall the eps sessions share the same GTP tunnel or mutiple gtp tunnel shall be created per eps session?

버너 2015-09-01 06:45:59

regarding to 1)

After UE is switched on, two default eps bearers are created and kept alive.

- Default eps session for internet APN which can be removed

if you enable WiFi instead of LTE data service at the UE

- Default eps session for IMS APN which is used

for initial / periodic registration to IMS (SIP REGISTER to CSCF)

and also for receiving call (waiting of SIP messages from IMS, e.g. SIP INVITE)

- Dedicated eps bearers for IMS APN to carry RTP will only be created and maintained

while VoLTE calling

regarding 2)

Sorry, I don't understand your question, "pdn connectivity request" is used for initial attachment.

If you are considering about creation of dedicated bearers for IMS APN, dedicated bearers are created only by IMS

- Initiated by CSCF -(Rx)-> PCRF -(Gx)-> PGW -(S5)-> SGW -(S11)-> MME -(S1)-> eNB/UE

regarding 3)

- Yes if "multiple PDN" is not used

(As I know "mutiple PDN" is defined but not serviced by any operators)

- A GTP tunnel can not be shared by several eps sessions

- One EPS bearer is made up of one GTP tunnel and several SDF templates

- Please see following articles

- https://www.netmanias.com/en/post/techdocs/5907/identification-identifier-lte/lte-identification-iii-eps-session-bearer-identifiers

- https://www.netmanias.com/en/?m=view&id=blog&no=5932

Sanjay 2016-09-27 20:12:49

Thanks a lot Netmanias, it is truely an excellent informative site.
I have one quiery, As we understood, we need a default bearer and dedicated bearer to IMS for the voice call, how many bearer we need for a normal internet usage ? As internet will be a separate PDN/APN.

Sanjay 2016-09-27 21:16:42

I got the answer for my query...I didnt read all the pages while asked this query..

sowjanya 2016-11-07 19:20:53

May I know the difference between EPS bearer QOS and subscriber's allowed QOS?

asfli asdf 2017-04-04 19:02:54

Thanks for your useful posts.

I had a question about "Default Bearer Establishment" (Section 3.1, Figure 3). In step 10 and 11 the MME and UE are informed about QoS parameters (such as QCI), however, the S-GW and eNodeB are not informed by anyone. As I know S-GW and eNodeB are envolved in policy enforcement process, so how they are informed about policy?

MANJULA R 2019-03-25 11:05:31

Hi,

very usefu posts,

I have a question regarding dedicated bearer , if there are more than one QCI 1 profile with different ARP values how PCRF picks the correct profile for a particular subscriber.

example if i have to prioritise some subscribers over others with same IMSI.

one way i undersatnd is to differentiate based on IMSI, but in case if its same IMSI and two profiles with different ARP values for QCI1.

manjunath UH 2019-12-20 17:40:30

Hi,

1) If UE have Multiple APN's, then whcih one select by PGW..

2) Diffrence b/w UE- AMBR & APN- AMBR

3) In whcih message PGW alocates IP to UE...

Tsvetan Filev 2020-04-29 00:58:38

Hi. Great article.

I have one quick question: How does the PGW (PCEF) bind the PCC rule to a given UE/bearer ?
Does it use UE IP from the SDF (DL/UL) and searches in the list of active bearers ?

prashant mundepi 2020-05-14 16:49:01

Hi, thanks for the great explanation !

is it possible to establish both the default and dedicated bearers during the initial attach? Its a requirement from an operator.

chandan dubey 2020-09-17 20:09:42

Hi Prashant,

Dedicated bearer is created only when condition is default must establish first ....

SCHIPPERS 2022-01-06 02:49:38

Hello, very usefull document,

I have a question and cannot understand when you mention "(UE IP, *, RTP, *, UDP)/(*, UE IP, RTP, *, UDP)":

What means the "*"? is this the Port number ?

And for the DL part shouldn't it be "(*, RTP, UE IP, *, UDP)" instead of "(*, UE IP, RTP, *, UDP)" ?

thank you very much !!

Umesh Patel 2022-02-28 18:05:11

Hello Netmanias,

Can you share details information about OCS (Onlice charging)

Thank you for visiting Netmanias! Please leave your comment if you have a question or suggestion.

Comment