Evaluating SIP Signaling Performance for VoIP over LTE Based ...

2015 9th International Conference on Next Generation Mobile Applications, Services and Technologies

Evaluating SIP Signaling Performance for VoIP over LTE based Mission-Critical Communication Systems Ashraf Ali, Mazin Alshamrani, Alhad Kuwadekar and Khalid Al-Begain University of South Wales, Cardiff, UK {ashraf.ali, mazin.alshamrani, alhad.kuwadekar, k.begain}@southwales.ac.uk

due to its ability to meet all the requirements for MCS. TETRA has short call-setup time (less than 500 ms) [7] which is of great importance for the MCS, it also provides bidirectional authentication between terminals and the core infrastructure in addition to air interface as well as end-to-end encryption. It supports Direct Mode of Operation (DMO) and Trunked Mode of Operation (TMO) that enable point to point and point to multi point communication in both modes that enhance the reliability and resilience for MCS. TETRA has a wide coverage due to the low frequency of the carriers; it also supports interoperability with other communication networks through gateways for wider accessibility domains and better interoperability between different groups. Finally, there are different levels of priorities that are defined in TETRA, pre-emptive call scheduling based on the priority class will guarantee that the more critical calls pass through the network with lower priority calls scheduled automatically at the bottom of the queue to be served afterwards. Despite the advantages and features of TETRA as MCS, it can only support up to 36Kbps in Release 1[7] and up to 115.2 Kbps in 25 KHz channel and 691.2 Kbps in 150 KHz channel in Release 2 [13]. The afforded bandwidth by TETRA for PPDR users is considered enough for voice and basic data services. However, due to the evolution of mobile communication technology which introduced to the users broadband services motivated by the increasing demand for more bandwidth driven by user’s needs, a gap between TETRA and commercial mobile technologies has been increasing in terms of capacity and broadband services supporting ability. PPDR users need services nowadays that need more bandwidth for their routine operations, and to be able to support the newly introduced Public Safety applications such as verification of biometric data, wireless video surveillance and remote monitoring, documents scan and database check, access to buildings blueprints, and remote emergency medical services [14]. This introduced the need for more MCS support for broadband bandwidth and accessibility for end users, which is already existing in nowadays broadband mobile technologies such as LTE in 4G communications. Using already existing LTE broadband mobile communication systems infrastructure for the conventional mobile communication along with the professional mission critical mobile services is what is considered the optimum migration from nowadays dedicated missioncritical systems. There are many challenges that need to be addressed for the new proposed system especially that LTE is not optimized for voice centric services that are mostly needed by PPDR member along with other data-

Abstract— Mission Critical Communications and Services are of a special nature that needs special requirements. Therefore, there are many challenges that face implementing mission critical communication system over commercial mobile communications networks. Signaling of SIP messages in the access technology domain and IMS domain needs enhancement to ensure meeting the mission critical service requirements. LTE, as a broadband technology that is widely deployed nowadays for providing 4G communication networking services, is facing a lot of challenges in terms of its ability to provide multimedia services. And due to the special nature of Mission Critical Systems (MCS) that have stringent requirements for SIP signaling specifically, the need for investigating the SIP signaling performance over LTE based MCS is of great importance. In this paper, the performance metrics for an SIP signaling over LTE-MCS will be defined, and the overall system performance measures were evaluated via discrete event simulator. KeywordsOPNET

Mission Critical; SIP; LTE; IMS; VoLTE;

I. INTRODUCTION Mission Critical Systems (MCS) are needed mainly as All-Time-Available backbone system for emergencies, crises, and disaster scenarios. It is used for Public Protection, and Disaster Relief (PPDR) operations done by law enforcement, emergency, and medical services teams. Due to the critical nature of MC systems, it has a set of strict requirements to ensure accomplishing tasks and duties that are strongly associated in most times to human lives and national security. Delay, Interoperability, Availability, Reliability, Security, and resilience are some of the requirements that are needed in MCS to ensure optimum functionality as required by MCS users. There are two main deployments of MCS; the first one is a dedicated system that it designed to meet all the MCS requirements and operating only for mission critical operations and tasks, and the second one is a commercial general purpose mobile communication system that is used for both conventional mobile communication between public mobile users and used also as MCS by professional team members, and finally there is a third hybrid approach that combines between the commercial and dedicated MCSs 0 . The dedicated mission-critical communication is already designed to meet the strict requirements of the mission critical services and communications. TErristrial Trunked Radio (TETRA) in Europe [7] and P25 0 in the USA are the most common standards being used nowadays for the mission-critical communications. TETRA nowadays is widely deployed in many countries 978-1-4799-8660-6/15 $31.00 © 2015 IEEE DOI 10.1109/NGMAST.2015.68

199

centric services. In this paper, we will focus on the delay of Session Initiation Protocol (SIP) signaling between different entities of the system to have better understanding of the LTE eligibility to be considered as MCS. The emerging of the LTE capabilities along with IP Multimedia Subsystems (IMS) based services such as Push over Cellular (PoC) requires further integration and optimization efforts in order to meet the strict requirement of MCS. In [15], the potential of GPRS and CDMA for mission critical use was analyzed by TETRA Association, the study shows that the latency of call setup time is not acceptable for MCS use. In [16] IMS was added on top of GPRS network to minimize the callsetup latency, which added advantages for group calling but with added call setup overhead. In [17] and [18], IMS with PoC on top of UMTS and CDMA2000 feasibility was investigated and shows that the overall system latency was reduced but still not enough for critical communications. In [19] and [20] the needs of the public safety community in HSPA and LTE was addressed. In this paper, the objective is to investigate more the SIP signaling over LTE to de-risk the compliance of any Mission Critical System for the End-to-End system performance targets. In addition to identifying the potential risks within the End-to-End system performance at the component performance level using simulation tools such as OPNET. The rest of this paper is organized as follows; Section II discusses LTE and the performance metrics that are related to SIP signaling. Section III, discusses the IMS and its influence on the end-to-end SIP performance, Section IV discusses the SIP protocol performance metrics and benchmarking challenges, Section V demonstrate the simulation scenarios and section VI will show the results along with discussion, and finally conclusions will be presented in Section VII. II.

the general architecture for LTE communication network. LTE is a strong candidate to be considered for the Mission Critical and Public Safety Communication Systems due to its ability to provide the end users with the broadband services capacity and other MCS requirements [3]. Delay and bandwidth are very important due to the fact that services require more bandwidth and less latency for sufficient QoS and reliable service. Clearly there are many challenges that the LTE faces in MCS deployment such as supporting Direct Mode Operation (DMO) between devices and providing the needed interfaces to ensure interoperability with already existing MCSs, in addition to supporting

Figure 1. System design LTE System Architecture

LTE SYSTEM AS MCS

group voice calls over IP packet switched structure. But at the same time the All-IP-Network flat architecture is considered a big advantage for LTE to overcome the interoperability and complexity issues that may emerge in the proposed project. Moreover, the spectral flexibility and the higher spectral efficiency allows for more bandwidth utilization and better resource allocation for the end users which enhances the QoS provided [4] and highlights LTE capabilities in terms of providing different QoS classes at different levels. LTE provides a set of service preferences for the user to meet a certain level of service requirements, and the requirements affect the scheduling and queueing user data, priority and preemption capabilities, and the access control treatment. All the aforementioned requirements are out of the scope of the bearer domain and managed by the access technology. At the access level, there are 16 priority classes that may be dedicated to public safety users to overpass the network overload access issues and at the network level, the Evolved Packet System (EPS) bearer, which is a logical channel between the UE and the P-GW at the far edge of the EPC, has two types; the first one is the Guaranteed Bit Rate (GBR) bearer where the user has a reserved resource during admission, and the Non-Guaranteed Bit Rate (None-GBR) bearer that is

The Long Term Evolution (LTE) standard was developed by the Third Generation Partnership Project (3GPP). The standardization started in Release 8 and continued until Release 10 for LTE-Advanced. It is considered a normal evolutional step of the mobile technologies but using revolutionary communication techniques at the physical layer to allow higher bandwidth and less latency for the end users. There are two parts of LTE Communication Network; the first one is the Radio Access Network (RAN) part which is known as Evolved UMTS Terrestrial Radio Access Network (E-UTRAN) that has enhanced performance compared with Universal Mobile Telecommunications System (UMTS) that was used in the 3G communication networks. Evolved UMTS (EUMTS) is mainly responsible for managing the whole radio stack signaling between the access point that is called evolved Node B (eNB) and User Equipment (UE). The other part is the core All-IP-Network part which is called the Evolved Packet Core (EPC) that is mainly responsible for managing the bearer services, mobility management, and interconnecting the interfaces and gateways with other domains and entities. Fig. 1 shows

200

based on best effort service without guarantees. There is also the Allocation ad Retention Priority (ARP) which determines the priority class of the connected bearer in addition to two other flags; the preemption capability flag, that determine if it is allowed to preempt another lower priority bearer. And the preemption vulnerability flag, which determines if the bearer may be preempted by another higher priority bearer. The APR will facilitate the decision of managing the bearer connection in overload conditions. In order to run MCS, the LTE needs to be coupled with the IP Multimedia Subsystem (IMS) to create an environment capable of supporting voice and video traffic in a shared packet data network. The next section speaks about IMS and its coupling with LTE. III.

users in addition to registration status record update. The Call Session Control Functions (CSCF) are responsible for handling the SIP signaling messages and packets in the IMS. The Proxy CSCF (P-CSCF) is the entry point into the IMS system. All SIP messages flow through the P-CSCF. The P-CSCF may also apply security or compression algorithms over the received traffic in addition to the quality of service control and bandwidth management. The interrogating CSCF (I-CSCF) is one of the main elements of the IMS systems. It is used during the registration process when the UE does not know which Serving CSCF (S-CSCF) should receive the request. The I-CSCF interrogates the HSS to obtain the address of the appropriate S-CSCF that should process the request. S-CSCF performs session control that has interfaces with the HSS to check and download the user profile information and also assign the Application Server (AS) for the user for further services in addition to enforcing the operator policy control. The SIP AS has an SIP interface with S-CSCF, and it is used to host specific IMS services. After the registration process is completed and the S-CSCF is allocated to the UE, the I-CSCF is no longer used for any further communication. All future communication happens between the UE, P-CSCF, and the S-CSCF. In this paper, the performance of the IMS system will not be investigated and will be left for future work. Hence, the focus of this paper will be on the SIP signaling performance for LTE without investigating the internal interfaces of IMS.

IP MULTIMEDIA SUBSYSTEM

IP Multimedia Subsystem (IMS) is designed and standardized by 3GPP [5] for providing multimedia services over mobile communication technologies beyond GSM. IMS is used for delivering the IP multimedia services between users and service providers. It gained its importance as an architectural framework for multimedia communications. SIP is the primarily signaling protocol used within IMS to create, modify and destroy multimedia sessions. Fig. 2 shows the IMS, architecture model. As defined by the standard, the IMS is operating as an interface between the service/application layer and the transport layer which enables the service providers and operator to control the user QoS based on its subscription profile. Moreover, it

IV. SIP PERFORMANCE METRICS There are mainly two types of protocols; the control plane set of protocols and the data plane protocols. The Session Initiation Protocol (SIP) is one of the control domain protocols and was fully standardized and specified by Internet Engineering Task Force (IETF) in RFC 2543 for the first version SIP 1.0 and in RFC 3261 for the second version SIP 2.0 [6]. SIP operates over IP protocol and considered as a communication protocol for signaling of real-time multimedia services such as voice and video and non-real-time services such as text messages and presence notifications. The protocol, which is text-based, mainly defines the signaling order between end users for call initiation, termination, in addition to modifying the call setup instantly during the call. It is also used for registering the users before the call being initiated. In this paper, the SIP message headers and signaling details will not be presented. But the performance issues and the challenge of enhancing SIP services performance will be highlighted and briefly discussed. Fig. 3 shows the signaling diagram for registering users, initiation and termination of a call between the caller, callee, and back to back SIP server. SIP performance has tremendous impact over the whole communication system due to the fact that it is responsible for initiation, managing, and termination of calls that embed real-time service exchange between end users. And due to the real-time services constraints especially with time limits, it is important to target all the performance issues of SIP to improve the overall system performance and to decide

Figure 2. IMS System Architecture

works as a hub point for all the SIP signaling that need to take place before, during, and after the call. For this purpose, there are different functions that are connected by interfaces to ensure integrity. The user’s subscription-related information is stored in the Home Subscriber Server (HSS) which also perform the authentication and authorization function of 201

Agreement (SLA) indicators for best network resources utilization and best end user Quality of Experience (QoE). The main metrics defined in RFC 6076 are the Registration Request Delay (RRD), Session Request Delay (SRD), and Session Disconnect Delay (SDD). RRD is the time needed for the user to finish the registration process successfully. SRD is the time needed to get a reply from the server side regarding the requested call setup from the user side; it is counted for both successful and unsuccessful call requests. If the call requested was successfully set then the call setup time will be simply SSD in addition to acknowledgment sending time. The SDD is the time difference between sending BYE message from the user side and the time of receiving 200OK confirmation from the server. In this paper, the call setup delay will be used to measure the system performance due to its importance in real-time multimedia services in general and in missioncritical communications specifically. The QoE for SIPbased systems is enormously affected by the call setup value. Based on [10] the call setup time means value can be up to 800 ms. However for LTE-based Mission Critical Systems which are supposed to work as a replica for traditional dedicated Mission Critical Systems such as TETRA [7], the call setup time need to be within 500 ms delay.

what is considered to be accepted metric to be considered as a benchmark for any other proposed solution. The protocol related performance metrics need to be identified to determine the way SIP is utilizing the system resources and how to maximize it. Moreover, in addition to the architectural design challenges that need to be targeted to enhance the SIP performance. Some of the protocol-related metrics in addition to implementation related metrics is discussed in [8], it shows different set of tests that measures the processing time for SIP messages, memory allocation, thread performance, and call-setup time. The results show that the performance of the proxy server changes by varying SIP related parameters and thus affecting the number of calls by seconds that the proxy server ca handle at a time. It also shows that the performance of SIP related architectures, such as the IP Multimedia Subsystems (IMS) that will be presented later, is more affected due to the heavy dependence of SIP signaling and SIP messages structure compared with a simple Proxy/Registrar Server. Furthermore, it is important to note that the performance of SIP signaling is enormously affected by the delay at different stages of registration, call initiation, and call termination processes. And the performance of SIP signaling will also affect the QoS of the offered service. Hence, the need to define the metrics that identify the performance measure for SIP is crucial for evaluation and performance comparison purposes. IETF proposed the criteria for the end-to-end SIP performance measures in RFC 6076 [9]. Due to the lack

V. SIMULATION SETUP AND SCENARIOS In this research study, we considered OPNET Modeler as a simulation tool as it provides the required level of simulation capabilities to implement and model different multimedia applications over LTE. The system design that was implemented and investigated in this paper is shown in Fig. 4 based on the configuration parameters on Table I. The implementation of the LTE network system is based on a single (EPC) that serves two eNBs where each has four clients. The clients in eNB1 are making SIP-based VoIP calls with the clients in eNB2 through the EPC in a Normal distribution call generation system using a fixed length for all VoIP calls duration. The EPC then connected to the SIP server, which is supposed to reflect the performance of the PCSCF in the IMS, that manage the registration, call initiation and call termination processes using the SIP signaling system and only through the IP cloud. In this study we considered the implementations without any background traffic in the LTE system and IP cloud to check and study the actual performance level for SIPbased VoIP applications within a best effort environment which helps with the results accuracy. In addition, the LTE implementations in this research study have not considered any mobility issues for its clients along with all the implemented scenarios. The simulation implementations has considered four scenarios based on the introduced design at Fig. 4 and the simulation parameters in Table I. The first scenario represents the basic implementation for VoIP applications over LTE using a single pair of UEs between client A-1 in eNB 1 and client B-1 in eNB 2. This scenario examines the best case implementation of the assigned network system with only one single call at the time. The second scenario has an additional connection with multiple calls with another pair of UEs

Figure 3. SIP signaling flow and performance metrics.

of a SIP benchmarking numbers to define the baseline performance of SIP signaling, RFC 6076 defines the performance metrics for SIP in VoIP applications to provide Key Performance Indicators and Service Level 202

VI. IMPLEMENTATION RESULTS

(client A-2 and client B-2) added to the first scenario. The same thing with the third scenario where additional pairs of calls been added (client A-3 and client B-3). Finally, the fourth scenario has a fourth additional pair between client A-4 and B-4. This gradual increase in the pairs of SIP-based VoIP calls from the first to the fourth scenarios aimed to check the performance of the SIP signaling system over LTE based communications with additional VoIP calls between different clients. The highest load of VoIP calls is represented in the fourth scenario that consumes higher bandwidth over LTE where all clients in each eNB are calling one single client in the other eNB. Therefore, the results of these implemented scenarios will be compared and studied throughout the research study in terms of the performance for SIP signaling and efficiency for LTE system. TABLE I.

As the main considerations in this study is for the SIP signaling and LTE performance for mission critical systems, the results representation focused on the call setup time and related LTE performance metrics. The optimum number of initiated calls for each pair of calls is falling between 150 to 180 calls for 30 minutes of simulation time with the uniform based distribution system for calls initiation. Table II shows the number of rejected calls in the overall system for the four scenarios with the implemented normal based system. The number of rejected calls has increased with the increased number of initiated call pairs. For calls implemented from Caller A-1, the number of failed calls initiation processes had been increased with the increased number of call pairs with scenarios S2, S3, and S4, where the total initiated calls over all scenarios is 56 calls. This increase in the failed initiation processes results from different causes even related to the SIP servers’ performance or LTE system performance.

SIMULATION PARAMETERS IN OPNET

A. LTE Network System 4

Number of Simulations

1

Background Traffic

Number of nodes for 2 Number of eNB: each eNB: Antenna Gain for eNB Maximum 15dBi eNB: Transmission Power: - 200 eNB Receiver eNB Selection dBm Sensitivity: Threshold: B. Applications: SIP Based VoIP

4 0.5 W - 110 dBm

Call Duration

Caller

10 Sec

Maximum Simultaneo us Calls

SIP Server

Unlimited Call/ Second Calls Start Time Offset:

Node A Node B User Agent Voice (Caller/Callee) Codec: 1 call at time GSM between each pair 13 Kbps Normal (150 sec, 100 sec)

Calls Inter-repetition Time:

Normal (20 sec, 5 sec)

CALLS STATISTICS FROM SIMULATION RESULTS

SIP calls statistics for the Implemented Scenarios S1: S2: S3: Scenario 1Pair 2Pairs 3Pairs Number of Calls 45 95 152 Rejected in the overall system Number of Calls 56 56 56 Initiated from Caller A-1 Number of failed calls initiation for 27 30 38 calls from Caller A-1

0%

VoIP Calls (Unlimited)

Figure 4.

TABLE II.

Callee

S4: 4Pairs 218

56

34

A. Call Setup Performance The importance of studying the call setup time is to analyze the SIP signaling performance during the main SIP signaling stage over different call sessions. As long as the call setup time for the majority of initiated SIPbased calls were in the acceptable range, the performance of the SIP signaling system falls in its acceptable level [9][11]. Fig. 5 represents the average call setup time for

Average SIP Call Setup Time (Sec)

Number of EPC:

128

Simulation Seed Number 30 Minutes = 1800 Seconds

Simulation Duration:

System design and implementation for SIP-based VoIP applications over LTE network system in OPNET

Figure 5. System Average call setup time in milliseconds

203

shown in Fig. 7. The average LTE delays for three pairs of VoIP calls is from 2.4ms to 3.5ms, and between 2.5ms to 4.3ms with four pairs of calls. The longest delays mostly happen at the system startup time and come to the stability level later during the simulation time.

all successful VoIP calls for the four implemented scenarios. The results show that the scenario with only one pair of calls had the lowest average of call setup time from 46 to 47 ms and increased up to 48.5ms with the scenario of two pairs. With three pairs of VoIP calls, the average call setup time has increased from 47ms to 49.5ms. The longest call setup time registered with the fourth scenario in which four pairs of VoIP connections were active at the same time. These simultaneous calls affected the SIP signaling performance and increased the average delay to reach up to 50.5ms. In general, the call setup time for successfully initiated calls over all scenarios is still in its acceptable level regarding the performance of the SIP signaling system. This came as a result of the representation for the LTE network system that implemented in its best effort without any assigned delays or background applications. B. LTE Downlink Packets Dropped The LTE parameters of the implemented system have a direct effect on the performance of the running applications. Moreover, real-time applications could be enhanced in case related LTE system behavior been considered with the required level of performance. Therefore, the LTE nature most considered during the studies of SIP-based VoIP over LTE. The average of packets dropped begun with 1 to 3 packets/sec for the single pair scenario and increased up to 6 to 18 packets/sec for the scenario with four pairs as shown in Fig. 6. The downlink packets dropped of LTE system shown a relation with successful SIP sessions where as long as the number of calls increased as the percentage of

Figure 7. Average LTE Delays in second for Caller A-1 node

VII. CONCLUSION AND FUTURE WORK

Average Downlink Packets Dropped (Sec)

Based on the simulation results, it is clear that there is increasing delay in the call setup if LTE communication system was used. This delay is increasing with the number of served client, which indicate that the Delay requirement or the maximum number of users that can be served at a time may not meet the mission critical service requirements. Hence, the need for decreasing the gap of call setup delay for commercial broadband systems compared with other dedicated mission-critical communications systems is of great importance and considered one of the main challenges for missioncritical communications. This means that there is a need for a new mechanism with minimum possible overhead to minimize access delay by exploring the LTE and IMS domains in addition to the interfaces between LTE and IMS and the interface between LTE and User Element. Moreover, so far we have used a simulation with static mobile nodes, the positions of the nodes are fixed. Hence, there is no handoff added complexity for the nodes moving between two cell domains. And by mobility we don’t only mean moving nodes but also a dynamic topology that supports handoff mechanisms between the subscriber stations and different base stations. Hence, we are looking for testing different communication scenarios for an end to end connectivity over LTE communication system. For such dynamic topology, we intend to measure the overall performance of the system in terms of SIP signaling and data streaming delay.

Simulation Time (Seconds)

Figure 6. LTE downlink packets dropped during VoIP calls for calls generated by Caller A-1 in packets/second

packet dropped relatively increased. C. Delays in LTE System The LTE system delays in the transferred data between LTE components affect the performance for real-time applications. The average LTE delays with one and two pairs of VoIP calls is between 2ms to 2.7ms as

204

REFERENCES

[12] Project 25 Technology http://www.project25.org

Interest

Group

(PTIG),

[13] ETSI, "Terrestrial Trunked Radio (TETRA); TETRA Enhanced Data Service (TEDS); Air Interface Specification," ETSI TS 100 392-2 V3.1.1, Sept. 2006.

[1] Ferrús, R., Sallent, O., Baldini, G., & Goratti, L. (2013). LTE: The technology driver for future public safety communications. IEEE Communications Magazine, 51(10), 154–161.

[14] Baldini, G.; Karanasios, S.; Allen, D.; Vergari, F., "Survey of Wireless Communication Technologies for Public Safety," Communications Surveys & Tutorials, IEEE , vol.16, no.2, pp.619,641, Second Quarter 2014.

[2] Blom, R., de Bruin, P., Eman, J., Folke, M., Hannu, H., Naslund, M., Synnergren, P. (2008). Public Safety Communication using Commercial Cellular Technology. In NGMAST 2008.

[15] TETRA MoU Association. Push To Talk over Cellular (PoC) and Proffessional Mobile Radio (PMR), TETRA 2004.

[3] Doumi, T., Dolan, M. F., Tatesh, S., Casati, A., Tsirtsis, G., Anchan, K., & Flore, D. (2013). LTE for public safety networks. IEEE Communications Magazine, 51(2), 106–112.

[16] Sanjay Kanti Das. Feasibility study of IP Multimedia Subsystems (IMS) based Push To Talk over Cellular for Public Safety and Security Communications. Master’s Thesis Department of Electrical and Communication Engineering. Helsinki University of Technology (HUT), 2006.

[4] Simic, M. B. (2012). Feasibility of long term evolution (LTE) as technology for public safety. In 2012 20th Telecommunications Forum (TELFOR) (pp. 158–161). [5] 3GPP TS 23.228, “Service Requirements for the Internet Protocol (IP) Multimedia Core Network Subsystem (IMS), Stage 1.”

[17] Balachandran, K. Budka, K.C Chu, T.P. Doumi, T.L Kang, J.H. Mobile Responder Communication Networks for Public Safety. IEEE Communication Magazine, Jan 2006

[6] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, A., Peterson, J., Sparks, R., Handley, M., and E. Schooler, "SIP: Session Initiation Protocol", RFC 3261, June 2002.

[18] Balachandran, K. Budka, K.C Chu, T.P. Doumi, T.L Kang, J.H., R. Whinnery. Converged Wireless Network Architecture for Homeland Security. Military Communications Conference, IEEE MILCOM2005, Atlantic City, NJ, Oct 2005.

[7] ETSI (2011), ETSI EN 300 392 V3.5.1, Terrestrial Trunked Radio (TETRA), Voice plus Data (V+D);

[19] Rolf Blom, Peter de Bruin, Jesper Eman, Mats Folke, et al. Public Safety Communication Using Commercial Cellular Technology. The second International Conference on Next Generation Mobile Applications, Services and Technologies, 2008.

[8] Bell Labs Technical Journal, “On SIP performance” Wiley Subscription Services, Inc., A Wiley Company, 2004. [9] D. Malas and A. Morton, “Basic Telephony SIP End-to-End Performance Metrics,” Technical Report RFC 6076, Internet Engineering Task Force (IETF), 2011, URL: http://tools.ietf.org/html/rfc6076.

[20] IPWireless. LTE addressing the needs of the Public Safety Community. 3GPP RAN Workshop on Rel-12 and Onward RWS120030. June 2012.

[10] ITU-T TR Q-series supplements 51 signaling requirements for IPQoS (December 2004). [11] M. Voznak, and J. Rozhon, “SIP back to back user benchmarking,” In Wireless and Mobile Communications (ICWMC), 2010 6th International Conference on, pp. 92-96. IEEE, 2010.

205