Video Transmission Quality Improvement Under Multi-Hop Wireless Network Architecture

  • Chih-Ang Huang
  • Chih-Cheng Wei
  • Kawuu W. Lin
  • Chih-Heng Ke
Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 260)


As wireless local area networks have been widely deployed and smart devices are becoming common in human everyday life, the demanding for video applications, such as video conference, video on demand, and video games is still increasing. However, the bandwidth in wireless networks is limited and channel quality is varying. As a result, how to provide a better delivered video quality over wireless networks is challenging. Also, the current solutions to better video transmission are almost only considering single hop transmission. When those solutions are applied in multi-hop wireless networks, such as wireless mesh network or VANET, the delivered video quality may not be good as one hop transmission. Therefore, we propose a new mechanism to improve video transmission in multi-hop wireless networks. The new mechanism contains two methods. One is to privilege the packets that have already traversed more hops. The other is control the number of retransmission for video packets with different importance. Through NS2 simulations, the results show the effectiveness of proposed mechanism.


Multi-hop wireless networks Video transmission MPEG TTL 


Advances in wireless network technology with increased bandwidth and the benefits of unwired connectivity have overcome numerous inconveniences commonly experienced in cabled networking. For example, through wireless distribution systems (WDSs), a wireless network might continually extend its coverage by joining two or more WDS-capable nodes. The network architecture that enables data transmission by relaying packets from one node to the next between the sending and receiving end is termed a multi-hop wireless network. Most studies regarding multi-hop wireless networks have focused on increasing the volume or speed of data transmission [1, 2, 3, 4, 5, 6], but few have examining improving the transmission quality of videos. Therefore, in this paper, the transmission quality of video streams through IEEE 802.11 [7, 8] and a multi-hop wireless network is examined.

Related Work

In wireless local area network communications, the IEEE 802.11 protocol is employed to access a shared microwave band of the transmission medium. Because of this characteristic of multi-hop wireless networks, preceding nodes and subsequent nodes compete for access to the channel, as shown in Fig. 1. Packet collisions resulting from this competition might cause latency and additional back-off time, thereby continually reducing the network bandwidth. Because of the competition between nodes, multi-hop wireless networks provide less bandwidth and longer latencies than single-hop wireless networks do. In this paper [9], a novel node-based priority calculation mechanism is presented to ease competition by granting higher priorities of channel access to packets that are sent earlier than later ones. When packets must be transmitted to a destination through a node, packets that are sent earlier compete with subsequent packets for channel access. However, if subsequent packets obtain the channel first, the utility is minimal because they cannot start transmission before the earlier packets finish transmission. Therefore, by granting channel access to earlier packets, the latency can be reduced and the bandwidth increased. Therefore, regarding access to the communication channel, the method presented here assigns a higher priority to earlier packets. In conventional techniques, all packets have equal priority for packet transmission, and early packets might compete with those sent later for channel access. The method proposed in this study increases the priority of the packets after passing through each node to reduce the competition between packets. In other words, the transmission priority is determined by the distance to the receiving end: Packets with shorter travel distances are assigned higher priority. By specifying a smaller contention window, packets closer to the receiving end gain access to the channel more easily. Similarly, by specifying a larger contention window, packets farther from the receiving end have a lower priority of access to the channel, and gaining access to the channel becomes more difficult. In this manner, higher transmission quality and operational efficiency in multi-hop wireless networks are achieved.
Fig. 1

Network environment of multi-hop wireless network

The Proposed Method

In addition to existing IEEE 802.11 specifications, this paper proposes adding two new mechanisms to improve the quality of video transmission in multi-hop wireless networks. The two mechanisms are described as follows (Tables 1 and 2):
Table 1

TTL and retransmission mechanism algorithm

TTL mechanism algorithm

Retransmission mechanism algorithm

If (qlen>threshold)


    △TTL = default_TTL – current_TTL;

     prob = ((qmax-qlen)/qmax)*(△TTL/maxhops);







I frame packet

    set RetryLimit = normal;

P frame packet


       set RetryLimit = less;


       set RetryLimit = normal;

B frame packet


       set RetryLimit = min;


       set RetryLimit = normal;

Table 2

Description of mechanism parameters




Current total queue length


Threshold of network congestion


Default TTL value of packet


Current TTL value of the packet


Probability that the packet enters the queue


Total queue length


Total number of nodes


Total number of nodes the packet passes through


Randomly generated probability value (0.0–1.0)


Maximum number of packet retransmissions


Normal number of retransmissions


Fewer numbers of retransmissions


Minimum number of retransmissions

  • Time to Live (TTL) Mechanism

Time to live is employed to protect packets traveling through numerous hops. Having passed through multiple contentions and approaching the receiving end, such packets should be granted higher priority through protective measures to increase the opportunity to enter the queue and the success rate for transmission, otherwise the video transmission bandwidth is affected and transmission quality compromised.
  • Retransmission mechanism

Vital packets are protected by altering the number of retransmissions. Because the space for node queuing is limited, if lower-order packets are retransmitted continually, the packet queue might overload and lose the higher-order packets because of its inability to accommodate them in the queue. The loss of higher-order packets might have severe consequences on video decoding and transmission quality.

TTL Mechanism

TTL refers to the longest time that a packet can live in a network. The TTL value of each packet varies in different operating systems. TTL primarily works by assigning a value to the packet transmitted by the sending end; the value decreases by 1 each time the packet passes through a node. When the TTL value drops to zero, the packet is discarded by the node that receives it. Specifying a TTL value prevents packets from random continuous travel among various network nodes, which is caused by unpredicted factors, before reaching the intended destination; this avoids affecting the overall network bandwidth and quality.

This mechanism first determines the number of nodes the packet has traveled through, and based on an algorithm, it assigns a higher priority to packets that have traveled through more nodes (i.e., by assigning a smaller TTL value) so that these packets have a greater opportunity to enter the queue for transmission. By assigning varying weights to packets through TTL values, packets that have traveled more nodes have a greater possibility of being transmitted to the receiving end. This mechanism is based on the concept that in a wireless network environment, each packet must pass through at least one node to obtain a transmission opportunity before it is transmitted again. In a multi-hop wireless network, all packets must pass through multiple nodes before reaching the receiving end; in other words, the packets must compete for priority in each node to be successfully transmitted to the receiving end. Conversely, when the wireless network environment is congested, the ability of the packets to travel through multiple nodes and successfully arrive at the receiving end decreases, thereby reducing the quality of the video transmission. Therefore, the use of the TTL mechanism could protect packets that have traveled through numerous hops by prioritizing them in the transmission queue, thus increasing the chance of successful transmission and greatly enhancing the bandwidth and video transmission quality.

According to the TTL mechanism algorithm, when the queue length surpasses a set threshold of network congestion, the number of nodes that a packet has passed through is equal to the default TTL value minus the current value. The probability of the packet entering the queue is determined by two factors: (1) the ratio of vacant space to the total length of the queue; a smaller remaining space represents a smaller probability of the packet being granted entry to the queue; and (2) the ratio of the number of nodes the packet has passed through to the total number of nodes (△TTL). The more nodes that a packet passes through increase the probability of the packet entering the queue; conversely, fewer nodes reduce the probability of a packet entering the queue. In addition, a probability value (tmpx) is randomly generated, with a value ranging between 0.0 and 1.0. When the probability of entering the queue (prob) is greater than or equal to the tmpx, the packet enters the queue for subsequent transmission; otherwise, the packet is discarded.

Retransmission Mechanism

Both IEEE 802.11 Distributed Coordination Function (DCF) and IEEE 802.11e [10] Enhanced Distributed Channel Access (EDCA) consist of a retransmission mechanism that might appear in two forms: the time limit and retry limit. This study is based on the retry limit mechanism because after compression by MPEG-4, video streams are encoded into I-frames, P-frames, and B-frames, which are subsequently encapsulated into frame packets. During transmission, these frame packets may experience unsuccessful transmission because of collisions or other factors. Under such circumstances, the working node determines the number of packet retransmissions based on the retransmission mechanism. For example, suppose an I-frame packet has a retry limit set to four. In the event that the frame packet transmission fails, the working node competes for its right to transfer again and retransmits the packet once retransmission permission is granted. If the frame packet cannot be transmitted to the next node after four attempts, the working node would discard the frame packet, thereby protecting the priority rights of other frame packets.

The retransmission mechanism is primarily based on adjusting the number of packet retransmissions. When a packet transmission fails, the so-called retransmission mechanism activates; when a packet fails to be correctly transferred to the receiving end, the node initializes packet retransmissions. In the current IEEE 802.11 configuration, the number of retransmissions is fixed and a priority is not assigned based on packet importance. In the present study, the varying effects of frames with differing importance levels on video transmission are considered; therefore, different probabilities for retransmission are assigned to frame packets based on their priority to enhance the quality of video transmission. When the network is busy, frame packets with higher priority are assigned more retransmissions, and lower priority frame packets have fewer opportunities for retransmission. In this manner, the probability of losing frame packets with higher priority is significantly reduced, and the likelihood of successfully transmitting frame packets with higher priority is increased; thus, the overall quality of video transmission improves considerably. Because the size of the queuing buffer at each node is limited, if frames of varying priorities are provided the same level of protection, the node would be busy retransmitting low-priority frame packets, and its queue would have insufficient space for high-priority packets, thereby resulting in poor performance. Therefore, this study adjusted the priority level of various packets, and reduced the number of retransmission attempts for low-priority packets when the network is in congestive or highly competitive conditions. The packets are discarded once the maximum retransmission number is attained so that subsequent packets can enter the transmission queue. The probability of successfully transmitting packets is significantly improved, as is the transmission bandwidth and quality of video transmission.

According to the retransmission mechanism algorithm description, the number of packet retransmissions is adjusted based on the priority of various packets. Important I-frames are always assigned four retransmissions, regardless of the wireless network condition; for second-level P-frames, the number of retransmissions is set at two in busy wireless network conditions and at four in slower periods. For low-priority B-frames, the number of retransmissions is set at one in congested wireless network conditions and four in slower periods. Therefore, the proposed retransmission mechanism is activated only when the wireless network is busy to assign more retransmissions to the higher order packets, thereby increasing the possibilities of successful transmission for higher order packets and effectively improving the video transmission quality.

Results and Analysis of the Simulated Experiment

Settings for the Simulated Parameters

In this section, a simulated experiment is conducted using NS2 [11, 12] network simulation software. In the experiment, a multi-hop wireless network environment is simulated, which consists of four wireless network nodes. Suppose that the sequence of these four nodes is AP 1, AP 2, AP 3, and AP 4. Direct data transfer may occur between adjacent nodes, and non-adjacent nodes do not interfere with one another. In other words, if data are to be transferred from AP 1 to AP 4, they must first be transferred AP 2, then to AP 3, and finally to AP 4; if data are to be transferred from AP 1 to AP 3, they must be transferred to AP 2 first, and then transferred from AP 2 to AP 3. Data can be directly transferred from AP 1 to AP 2 because these two nodes are adjacent nodes and capable of direct data transfer. The simulative parameters of the experiment are set as shown in Table 3.
Table 3

Simulated parameter settings



Set value


Threshold of network congestion



Default TTL value



Total queue length at the node



Normal number of retransmissions



Fewer number of retransmissions



Minimum number of retransmissions


Results and Analysis of the Simulated Experiment

In the experiment, a data stream was added to AP 1, which will be transferred to AP 4; data streams are also added to AP 2 and AP 3, which act as sources of interference during data transfer. After the data are transferred, the video frames transmitted in this experiment are decoded. By analyzing the peak-to-signal noise ratio (PSNR) of the decoding process, the proposed mechanism can be assessed for its improvement of the transmission quality of MPEG-4 compressed videos.

This simulated experiment generates three sets of data. The first set of data is obtained through the process of video transmission using the existing IEEE 802.11 mechanism; the second set is obtained from the experiment that incorporates the TTL mechanism into the 802.11 mechanism; and the third set of data is obtained from the simulated experiment that incorporates both the TTL and retransmission mechanisms.

Table 4 presents the Foreman video data. A total of 400 frames were extracted from the source video, including 51 I-frames, 83 P-frames, and 266 B-frames. The statistics of the video frames transmitted through the IEEE 802.11 mechanism (Table 5) show that 21 I-frames were received by the receiving end, 30 I-frames were lost; 50 P-frames were received and 33 were lost; 144 B-frames were received and 132 were lost.
Table 4

Foreman video data



Number of frames

Total number of frames










Table 5

Statistics of video frames transmitted using (a) IEEE802.11, (b) TTL mechanism and (c) TTL and retransmission mechanism




TTL and retransmission











Total transmissions










Total receptions










Number of losses










The statistics of video frames transmitted by using the TTL mechanism (Table 5) show that after the TTL mechanism was incorporated, 31 I-frames were received by the receiving end, which is 47.6 % higher than the transmission with IEEE 802.11 mechanism. The number of P-frame transmissions increased to 55 frames from the previous 50 frames using IEEE 802.11, which is a 10 % increase. Finally, 168 frames of B-frame packets were received compared to the 134 frames transmitted by using IEEE 802.11, indicating a 25.4 % increase.

The statistics of video frames transmitted using TTL and the retransmission mechanism (Table 5) show that after the TTL and retransmission mechanism was used, 35 I-frames were received by the receiving end, a 66.7 % increase over the transmission using the IEEE 802.11 mechanism; the number of P-frame transmissions increased to 58 from the 50 frames transmitted using IEEE 802.11, showing a 16 % increase; finally, 154 B-frame packets were received. Only 134 frames were received when using IEEE 802.11, indicating a 14.9 % increase.

The results of the average PSNR presented in Table 6 show that the average PSNR obtained by using IEEE 802.11 was 21.29 after the discarded frames were counted and compared to the source data. After incorporating the TTL mechanism, the transmission in the same simulation environment generated an average PSNR of 23.363; a 9.7 % improvement over IEEE 802.11. After the TTL and retransmission mechanism was added, the average PSNR was raised to 23.518 under the same experimental conditions, indicating a 10.5 % improvement over the IEEE 802.11 mechanism.
Table 6

Average PSNR of the foreman video


IEEE 802.11

TTL mechanism

TTL and retransmission mechanism

Average PSNR




Overall, the TTL and retransmission mechanism proposed in this study can effectively improve the quality of video transmission.

Conclusion and Research Suggestions

This study proposed two mechanisms for improving the quality of video transmission in multi-hop wireless networks: the queuing mechanism implemented by adjusting packets based on TTL analyses, and adjustment of retransmissions based on the varying priorities of differing packets. The analytical results of a simulated experiment conducted on an NS2 network confirmed that, compared to the traditional IEEE 802.11 mechanism, the two mechanisms proposed in this study can effectively improve the quality of video transmission in multi-hop wireless networks.

In this study, fixed nodes were employed in a multi-hop wireless network. Future studies should simulate actual scenarios occurring in multi-hop wireless networks, such as environments in which multiple nodes randomly move within a certain range, or in which specific nodes remain static but others move.


  1. 1.
    Shugong X, Saadawi T (2001) Does the IEEE 802.11 MAC protocol work well in multi-hop wireless ad hoc networks? IEEE Commun Mag 39(6):130–137CrossRefGoogle Scholar
  2. 2.
    G. Bianchi (2001) Performance analysis of the IEEE 802.11 distributed coordination function. pp 535–547Google Scholar
  3. 3.
    Hsieh H-Y, Sivakumar R (2002) IEEE 802.11 over multi-hop wireless networks: problems and new perspectives. In: IEEE 56th VTC 2002Google Scholar
  4. 4.
    Hsieh H-Y, Sivakumar R (2001) Improving throughput and fairness in multi-hop wireless networks. In: Proceedings of ICN, Colmar, France, July 2001Google Scholar
  5. 5.
    Nguyen LT, Beuran R, Shinoda Y (2007) Performance analysis of IEEE 802.11 in multi-hop wireless networks. In: Proceedings of the 3rd international conference on Mobile ad-hoc and sensor networks, ser. MSN’07. Springer, Berlin, Heidelberg, pp 326–337Google Scholar
  6. 6.
    Mustapha I, Jiya JD, Monguno K (2010) Throughput analysis of IEEE 802.11 MAC protocol in multi-hop wireless Ad-Hoc network. J Sci Technol Res 9(1):113–122Google Scholar
  7. 7.
    IEEE (1999) Wireless LAN media access control (MAC) and physical layer (PHY) specifications. IEEE Std. 802.11Google Scholar
  8. 8.
    Crow BP, Widjaja I, Kim LG, Sakai PT (1997) IEEE 802.11 wireless local area networks. IEEE Commun Mag 35(9):116–126Google Scholar
  9. 9.
    Lee SH, Yoo C (2010) Contention avoidance with hop based priority in 802.11e multi-hop network. Consumer electronics (ICCE), 2010 digest of technical papers international conference, January 2010, pp 151–152Google Scholar
  10. 10.
    IEEE Std 802.11e-2005 (2005) Wireless LAN medium access control (MAC) and physical layer (PHY) specifications amendment 8: medium access control (MAC) quality of service enhancementsGoogle Scholar
  11. 11.
    The network simulator ns-2 [Online]. Available:
  12. 12.
    Evaluation of video stream quality over IEEE 802.11e EDCF [Online]. Available:

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  • Chih-Ang Huang
    • 1
  • Chih-Cheng Wei
    • 1
  • Kawuu W. Lin
    • 1
  • Chih-Heng Ke
    • 2
  1. 1.Department of Computer Science and Information EngineeringNational Kaohsiung University of Applied SciencesKaohsiungTaiwan, Republic of China
  2. 2.Department of Computer Science and Information EngineeringNational Quemoy UniversityKinmenTaiwan, Republic of China

Personalised recommendations