N10-009 Question 8
Single answerLayer 4 - TransportA company has deployed a voice-over-IP (VoIP) softphone application to remote users. After a firewall policy change, users can still log in to the VoIP service, but during calls they experience choppy audio and frequent dropouts. Packet captures show the signaling session establishes successfully, but the voice stream suffers when the network is congested. Which transport-layer change would MOST likely improve call quality while keeping the application behavior appropriate for real-time voice traffic?
- A
Configure the voice media stream to use UDP instead of TCP
- B
Force the softphone signaling and media traffic to use only TCP port 443
- C
Enable larger TCP receive windows for the voice media stream
- D
Require the voice application to retransmit lost audio packets until acknowledged
Show answer and explanation
Correct answer: A
Explanation
This question tests practical understanding of Layer 4 transport behavior. For real-time applications such as VoIP, the transport protocol choice directly affects latency, jitter, and packet recovery behavior. UDP is commonly used for voice media because it is connectionless and does not provide retransmission, sequencing enforcement, or acknowledgments like TCP. That makes it better suited for time-sensitive traffic where delayed packets degrade user experience. TCP is more appropriate for applications that require reliable, ordered delivery, such as file transfers or web content. In common VoIP designs, signaling may use TCP or UDP depending on the protocol and implementation, but the media path is typically carried over UDP. This aligns with IETF real-time media practices, including RTP running over UDP, and with Network+ best-practice distinctions between reliable transport and low-latency transport.
- A. Correct.
Correct. Real-time voice media is typically carried over UDP because UDP has lower overhead and does not wait for acknowledgments or retransmissions. For voice traffic, late packets are often as useless as lost packets, so avoiding TCP-style recovery behavior helps reduce latency and jitter. This is the transport-layer choice most aligned with maintaining call quality during congestion.
- B. Incorrect.
Incorrect. Although tunneling traffic over TCP 443 may help with firewall traversal, using TCP for real-time voice media can worsen performance because retransmissions, in-order delivery requirements, and congestion recovery can increase delay and jitter. The scenario already indicates signaling works; the issue is media quality under congestion.
- C. Incorrect.
Incorrect. TCP receive window tuning is relevant to TCP flow control, but it does not solve the core issue for real-time audio. Larger windows can improve throughput for bulk transfers, not interactive media that is sensitive to delay variation. Voice streams generally prioritize timely delivery over reliable retransmission.
- D. Incorrect.
Incorrect. Requiring retransmission and acknowledgment of lost audio packets introduces additional delay, which is harmful to conversational voice. This reflects a common misconception that reliability is always preferable. At the transport layer, real-time applications usually accept some packet loss rather than incur the latency caused by recovery mechanisms.