LTE Protocol Quick Guide – LTE S1 AP and X2 / LTE NAS : PLMN Identity Encoding

Posted on by
Reply

Source : X2AP – 36.423.960 : 9.2.4   :   S1AP 36.413.970 9.2.38 : NAS 24.008-9a0  10.5.1.13

Length : 3 Bytes

 - digits 0 to 9, encoded 0000 to 1001,

- 1111 used as filler digit,two digits per octet,

- bits 4 to 1 of octet n encoding digit 2n-1

- bits 8 to 5 of octet n encoding digit 2n

 -The Selected PLMN identity consists of 3 digits from MCC   followed by either 

-a filler digit plus 2 digits from MNC (in case of 2 digit MNC) or
-3 digits from MNC (in case of a 3 digit MNC)

Examples :

MCC=123 and MNC=456  will have a PLMN Identity = 214365

MCC=234 and MNC=05 will have a PLMN Identity = 3204f5

NAS

PLMN List

PLMN1 = MCC Digit2,MCC Digit1,MNC Digit 3,MCC digit3,MNC digit2,MNC digit1

NAS Example : MCC=123 MNC=456  :: PLMN ID=216354

 

Ideal LTE/UMTS/GSM Protocol Test Tool Selection Tips : For Feature Testing or Performance Testing of Network Elements

Posted on by
Reply

Currently network OEM’s/Operators/Integrators/Training-Providers, looking for a new test solution, either choose from an off-the-shelf test tool or write their own bespoke solution internally. Here are some quick tips to consider when choosing a protocol test tool…

Test Tool Selection Criteria :

Capital Cost: Sometimes cost becomes the sole deciding factor. But more often than not cost takes a back seat if there is not much of a choice of tools to choose from.

Running/Operating Cost: This is probably the most important factor.  You have to take into account the overall running costs by considering the man hours, licenses, support costs, etc, over the lifetime of the test program.

Flexibility: In protocol testing, flexibility would mainly refer to the capability of being able to introduce error legs or edit protocol messages without much effort.  Beware though: flexibility does come at a cost. In most cases it means that the back end is much more complex and difficult to maintain.

Usability: What is the point of a tool if users can’t use it? Or if it’s so difficult to use that they avoid it?

Extensibility: Is the test tool future proof ? Does it have the right architecture to support future features and scenarios without much architectural changes.

Maintainability: How much effort and resources does the tool require to maintain and run? Is it easy, part-time job or will you need a whole team of developers/testers to maintain the tool itself.

Proven:  Is there a track record for the tool and its functionality? Are current users happy with it?

Support: There will always be a need for support.  How good is the support for this tool? Is it quick, friendly and effective or does it feel like banging your head against a wall?

General: As a rule thumb the degree of extensibility is directly proportional  to flexibility.  Flexibility is inversely proportional to usability.  It is very difficult to find a  test solution that fits all the facets.*

What do you think? All comments welcome

 

*Ahem… sorry for the plug but, yes, I admit it: I believe Mantra 3000  is the best of the bunch.

Mantra Logfile : LTE S1AP Initial Attach with IMSI and then Reattach with GUTI

Posted on by
Reply

Log Name: MantraLog_AttachIMSI_UserPlane_Detach_ReAttachGUTI_UserPlane_GUTI.mlog.pcap

Log Source:  Test Tool : Mantra 3000  OS: CentOS 6.2  Hardware: PC intel i3 3GB RAM

Notes: Simulated Elements: 1 x MME, 2 x eNodeB, 1xSGW , Exposed Interfaces: S1-MME, S1U

Scenarios:  IMSI_Attach—->Default eRAB—->Detach—->GUTI_Attach—->Detach

To download the log file please just fill out the form below:


Your Name (required)

Your Email (required)

Top 10 tips for testing real UEs against LTE eNodeB

Posted on by
65

Objective: To test real LTE eNodeBs against Real UEs/MEs with minimum investment. You have an eNodeB , some real UE’s and  a list of test-cases  which demand the earth! Setting up a lab with real EPC elements is extremely expensive and is also inflexible  when it comes to testing all those borderline test-cases / alternative scenarios.

So how does one start?

1. Firstly, select the Type of UEs/MEs you want to test with. Back in the UMTS days this used to be a very difficult task as there were a plethora of devices to choose from such as Samsung, Nokia, Apple, HTC, Novatel, Sony-Ericsson and many others. For LTE this list is much smaller and quite consistent as almost all the UE’s are based upon the Qualcomm chipset – be it Samsung  or Apple.

2. Select the Frequency Band for your UEs. This should basically match the  frequencies that can be set on the eNodeB. A good introduction to the various Frequency bands are available at http://en.wikipedia.org/wiki/E-UTRA#Frequency_bands_and_channel_bandwidths

3. UE Automation is another factor which may narrow down your list of UE’s to use .  Almost all the present day UE’s are smartphones. Therefore developing an Android or iOS application will help automate the calls as well as generate statistics.                           However, be aware that  developing applications for iOS is not free and requires a membership (around 99 USD per year) . There is also a lot else to consider before you can develop an automation App and is something that should be thought about. If you are aiming at a high end system  for automated UE/ME testing then SITE  or IXIA or Dyaptive would fit the bill.

4. How many UEs do you want to test with? If it is just one or two then setting up the lab in an open environment should work fine. But if  the number runs into the tens and hundreds then one must consider buying RF shielded cages/anachoic chambers along with RF attenuators. Currently there are several automatic RF attenuators in the market which allow you to control the attenuation with granularity as low as 0.1 dB.

5.  You will also need to buy some SIM (USIMs) cards which can be used as Test SIMs. The cost of Test USIMs is pretty low, and a set of 50 Test USIMs can be purchased for around 500 USD. You can buy Test USIM’s from Comprion or Gemalto .

6. RF Radiation: One must be aware that FCC guidelines restrict transmission of any RF without the proper permits or license. The Carriers and Operators like AT&T , Verizon etc pay hundred’s of million dollars to  buy the licenses. Obviously they wouldn’t want you to be radiating with their frequency ranges. The only option is to set up the eNodeB in as shielded an environment as possible.  Besides this  one also must make sure that the eNodeb radiates at the minimum power possible along with using attenuators  as well as wired connections.

6. Master the eNodeB MMI/OMC/LMT . It is very important to get a basic understanding of the eNodeB MMI/OMC.  You would want the capability to set the following at least : RF Sectors, Frequency/Band, Power Settings  IP Addresses, SCTP Settings, IPSEC capability etc.

7. Choose an EPC Simulator. There are several options that are out on the market. But one must choose a simulator/emulator which allows them flexibility and usability at the same time.  Be wary that most the solutions out there can only do either.  Capability wise you need to choose a simulator which can do at least 400-500 UE’s/ME’s simultaneously. (Full disclosure: we think our very own EPC simulator – Mantra 3000 – is the best on the market for all these reasons and more…).

8. Select an EPC Simulator that can do a Direct Gi Connection.  This way you could test real internet traffic testing with the UE’s/ME’s.

9. Performance wise you need to select a simulator which can do at least wire rate traffic on a 10G interface. Hardware wise you need to select a simulator which can run on any commercially available hardware so that you  set the price and you set the performance level.

10. Lastly you need to select an EPC simulator that allows you to  easily configure at least  the following without changing any scripts:

  • MME Code
  • MME Group, 
  • PLMN Identity, 
  • IMSI, 
  • UE Authentication Parameters like K Value,  
  • Message timeout parameters, 
  • SCTP  parameters/timers.
  • Failure Control Flags
  • Different  legs for alternative scenarios