๋ณธ๋ฌธ ๋ฐ”๋กœ๊ฐ€๊ธฐ

๐“ก๐“ธ๐“ธ๐“ถ5: ๐’ฆ๐‘œ๐“‡๐‘’๐’ถ ๐’ฐ๐“ƒ๐’พ๐“‹/Computer Network(COSE342)

[์ปดํ“จํ„ฐ๋„คํŠธ์›Œํฌ] CH2. Application Layer (#4)

P2P

1. File distribution time

ํ•œ ์„œ๋ฒ„์—์„œ N๊ฐœ์˜ peer๋กœ size F์ธ ํŒŒ์ผ์„ ๋ณด๋‚ด๋Š”๋ฐ ์–ผ๋งˆ๋‚˜ ๊ฑธ๋ฆด๊นŒ??

(Peer์˜ upload/download capacity๋Š” ํ•œ์ •๋˜์–ด ์žˆ๋‹ค)

1. Client-Server

  1. Server transmission : ํŒŒ์ผ์„ N๊ฐœ๋กœ ๋ณต์‚ฌํ•ด์„œ ์ˆœ์ฐจ์ ์œผ๋กœ ๋ณด๋‚ด์•ผ ํ•จ
    • ๋ณต์‚ฌ๋ณธ ํ•˜๋‚˜ ๋ณด๋‚ด๋Š”๋ฐ ๊ฑธ๋ฆฌ๋Š” ์‹œ๊ฐ„ = F/us (์ „์ฒด์šฉ๋Ÿ‰ / ์‹œ๊ฐ„๋‹น upload ํ•  ์ˆ˜ ์žˆ๋Š” capacity)
    • N๊ฐœ ๋ณด๋‚ด๋Š”๋ฐ ๊ฑธ๋ฆฌ๋Š” ์‹œ๊ฐ„ = NF/us
  2. Client : ๊ฐ client๋Š” ํŒŒ์ผ์„ ๋‹ค์šด๋กœ๋“œ ๋ฐ›์•„์•ผ ํ•จ
    • d_min = download rate์˜ ์ตœ์†Ÿ๊ฐ’ (์‹œ๊ฐ„ ๋‹น ๋ฐ›์„ ์ˆ˜ ์žˆ๋Š” ํŒŒ์ผ (๊ฐ€์žฅ capacity๊ฐ€ ์ž‘์€ ๊ฒƒ ๊ธฐ์ค€))
    • client๊ฐ€ ๋‹ค์šด๋กœ๋“œ ํ•˜๋Š” ์‹œ๊ฐ„์˜ ์ตœ์†Ÿ๊ฐ’ = F/d_min

์ฆ‰ client-server approach๋ฅผ ์‚ฌ์šฉํ•ด์„œ F ๋งŒํผ์˜ ํŒŒ์ผ์„ N clients์—๊ฒŒ ๋ถ„๋ฐฐํ•˜๋Š”๋ฐ ๊ฑธ๋ฆฌ๋Š” total delay

D_c-s >= max {NF/us, F/d_min}

2. P2P

  1. Server transmission : ์ตœ์†Œํ•œ ํ•˜๋‚˜์˜ copy๋ฅผ upload ํ•ด์•ผ ํ•จ
    • ํ•œ ๊ฐœ์˜ copy ๋ณด๋‚ด๋Š”๋ฐ ๊ฑธ๋ฆฌ๋Š” ์‹œ๊ฐ„ = F/us
  2. Client : ๊ฐ client๊ฐ€ copy๋ณธ์„ ๋‹ค์šด๋กœ๋“œ ํ•ด์•ผํ•จ
    • Min client download time = F/d_min
  3. Clients : ์ข…ํ•ฉํ•ด์„œ ์ด NF bits๋ฅผ ๋‹ค์šด๋กœ๋“œ ํ•ด์•ผํ•จ
    • ์ตœ๋Œ€ ์—…๋กœ๋“œ rate (์ตœ๋Œ€ ๋‹ค์šด๋กœ๋“œ ์†๋„๋ฅผ ์ œํ•œํ–ˆ์„ ๋•Œ) = us+∑ui

์ฆ‰, server-client๋Š” N๊ฐœ์˜ ํŒŒ์ผ์„ ์˜ฌ๋ฆด ๋•Œ, ์„œ๋ฒ„๊ฐ€ ์ˆœ์ฐจ์ ์œผ๋กœ(capacity๋งŒํผ์”ฉ) ๋‹ค ์˜ฌ๋ ค์•ผํ•˜๊ณ , P2P์—์„œ๋Š” peers๋“ค์ด 1๊ฐœ ์ด์ƒ์”ฉ ์˜ฌ๋ ค์„œ N๊ฐœ๊ฐ€ ๋˜๋ฉด ๋˜๋‹ˆ๊นŒ, P2P์˜ file distribution time์ด ๋” ์งง๋‹ค.

2. A case study : BitTorrent

P2P File Distribution : BitTorrent

  • ํŒŒ์ผ์„ 256kb์˜ ๋ฉ์–ด๋ฆฌ๋กœ ์ชผ๊ฐ ๋‹ค
  • torrent์— ์žˆ๋Š” peer๋“ค์ด file ๋ฉ์–ด๋ฆฌ๋ฅผ ๋ณด๋‚ด๊ณ  ๋ฐ›๋Š”๋‹ค.
  • Tracker : torrent์— ์ฐธ์—ฌํ•˜๋Š” peer๋“ค์„ ์ถ”์ ํ•œ๋‹ค
  • Torrent : ํŒŒ์ผ ๋ฉ์–ด๋ฆฌ๋ฅผ ๊ตํ™˜ํ•˜๋Š” peer๋“ค์˜ ๊ทธ๋ฃน
  • torrent์— ์ฐธ์—ฌํ•œ peer
    • ์ฒ˜์Œ์—” ์•„๋ฌด๊ฒƒ๋„ ์—†์ง€๋งŒ ์ ์ฐจ ์‹œ๊ฐ„์ด ์ง€๋‚  ์ˆ˜๋ก ๋‹ค๋ฅธ peers์˜ chunk๊ฐ€ ์ถ•์ ๋œ๋‹ค
    • tracker์— ๋“ฑ๋กํ•ด์„œ peer์˜ ๋ชฉ๋ก์„ ๊ฐ€์ ธ์˜ค๊ณ , peers("neighbors")์˜ ํ•˜์œ„ ์ง‘ํ•ฉ์— ์—ฐ๊ฒฐํ•œ๋‹ค.
  • ๋‹ค์šด๋กœ๋“œ ํ•˜๋Š” ๋™์•ˆ peer๋Š” chunks๋ฅผ ๋‹ค๋ฅธ peers์—๊ฒŒ ์—…๋กœ๋“œํ•œ๋‹ค.
    • peer๋Š” chunk๋ฅผ ๊ตํ™˜ํ•˜๋Š” peers๋ฅผ ๋ฐ”๊ฟ€ ์ˆ˜ ์žˆ๋‹ค.
    • ๋งŒ์•ฝ peer๊ฐ€ ์ „์ฒด ํŒŒ์ผ์„ ๊ฐ–๊ฒŒ ๋˜๋ฉด, ๊ทธ๊ฒƒ์€ ๋– ๋‚  ์ˆ˜๋„, torrent์— ๋‚จ์•„ ์žˆ์„ ์ˆ˜๋„ ์žˆ๋‹ค.

Exchanging File Chunk

Requesting chunks

  • ์–ด๋Š ์‹œ๊ฐ„์—๋‚˜, ๋‹ค๋ฅธ peer๋Š” ํŒŒ์ผ chunk์˜ ๋‹ค๋ฅธ ๋ถ€๋ถ„์„ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค.
  • ์ฃผ๊ธฐ์ ์œผ๋กœ alice๋Š” ๊ฐ๊ฐ์˜ peer์—๊ฒŒ ๊ทธ๋“ค์ด ๊ฐ€์ง€๊ณ  ์žˆ๋Š” chunk ๋ชฉ๋ก์„ ์š”์ฒญํ•œ๋‹ค.
  • alice๋Š” ๊ฐ€์žฅ ๋“œ๋ฌธ(?) missing chunk๋ฅผ peers์—๊ฒŒ ์š”์ฒญํ•œ๋‹ค.

Sending chunks : tit-for-tat

  • Alice๋Š” 4๋ช…์˜ peers์—๊ฒŒ ๊ฐ€์žฅ ๋น ๋ฅธ ์†๋„๋กœ ๊ทธ๋…€์˜ chunk๋ฅผ ์ „์†กํ•œ๋‹ค.
    • ๋‹ค๋ฅธ peers๋Š” Alice์—๊ฒŒ ์งˆ์‹ํ•œ๋‹ค(choked) ; ๊ทธ๋…€๋กœ๋ถ€ํ„ฐ chunks๋ฅผ ๋ฐ›์ง€ ๋ชปํ•จ
    • 10์ดˆ๋งˆ๋‹ค ์ƒ์œ„ 4๊ฐœ๋ฅผ ์žฌํ‰๊ฐ€ํ•œ๋‹ค.
  • 30์ดˆ๋งˆ๋‹ค ๋ฌด์ž‘์œ„๋กœ ๋‹ค๋ฅธ peer์„ ๊ณจ๋ผ์„œ chunks๋ฅผ ๋ณด๋‚ธ๋‹ค
    • Optimistically unchoke this peer
    • ์ƒˆ๋กœ ์„ ํƒํ•œ peer๊ฐ€ top 4์— join ํ•  ์ˆ˜ ์žˆ๋‹ค.

Streaming and CDN

Video Streaming and CDNs

  • video traffic : ์ธํ„ฐ๋„ท bandwidth์˜ ๊ฐ€์žฅ ์ฃผ์š” consumer! ํŠธ๋ ˆํ”ฝ์˜ ๋งŽ์€ ๋ถ€๋ถ„์„ ์ฐจ์ง€
    • Netflix, Youtube๋Š” downstream residental ISP Traffic์˜ ๊ฐ๊ฐ 37%, 16%๋ฅผ ์ฐจ์ง€ํ•จ
    • ~1B๋Š” ์œ ํŠœ๋ธŒ ์œ ์ €, ~75M ๋Š” ๋„ทํ”Œ๋ฆญ์Šค ์œ ์ €
  • Challenge
    • scalability (ํ™•์žฅ์„ฑ)
      : ์–ด๋–ป๊ฒŒ ~1B ์œ ์ €๋“ค์—๊ฒŒ ๋„๋‹ฌํ•  ๊ฒƒ์ธ๊ฐ€.
      ํ•˜๋‚˜์˜ mega-video server๋กœ๋Š” ์ž‘๋™ํ•˜์ง€ ์•Š์„ ๊ฒƒ
    • heterogeneity (์ด์งˆ์„ฑ)
      : ์‚ฌ์šฉ์ž๋“ค๋งˆ๋‹ค capabilities์ด ๋‹ค๋ฅด๋‹ค.
      ex. ์œ ์„  vs ๋ชจ๋ฐ”์ผ, bandwidth๊ฐ€ ํ’๋ถ€ํ•œ ๊ณณ๊ณผ ๊ทธ๋ ‡์ง€ ์•Š์€ ๊ณณ
  • Solution : distributed, application-level infrastructure!!!

1. Streaming multimedia

Multimedia : Video

  • Video๋ž€ : ์ผ์ •ํ•œ ์†๋„๋กœ ํ‘œ์‹œ๋˜๋Š” ์ด๋ฏธ์ง€๋“ค์˜ sequence
  • Digital image : ํ”ฝ์…€๋“ค์˜ ๋ฐฐ์—ด. ๊ฐ ํ”ฝ์…€๋“ค์€ bits๋กœ ํ‘œ์‹œ๋จ
    -> CBR (=Constant Bit Rate)
  • Encoding : ์ด๋ฏธ์ง€ ๋‚ด๋ถ€์™€ ์ด๋ฏธ์ง€ ์‚ฌ์ด์˜ ์ค‘๋ณต์„ฑ์„ ์ด์šฉํ•ด ์ด๋ฏธ์ง€๋ฅผ ์ธ์ฝ”๋“œํ•˜๋Š”๋ฐ ์‚ฌ์šฉ๋˜๋Š” ๋น„ํŠธ์ˆ˜๋ฅผ ์ค„์ธ๋‹ค
    • spatial : ์ด๋ฏธ์ง€ ์•ˆ์—์„œ ์”€
    • temporal : ํ•œ ์ด๋ฏธ์ง€์—์„œ ๋‹ค์Œ ์ด๋ฏธ์ง€๋กœ ๋„˜์–ด๊ฐˆ ๋•Œ ์”€
      -> VBR (=Variable Bit Rate) : ๋น„๋””์˜ค์˜ ์ธ์ฝ”๋”ฉ ์†๋„๋Š” spatial, temporal coding changes ์–‘์— ๋”ฐ๋ผ์„œ ๋ณ€ํ•œ๋‹ค.
  • spatial coding example : N๊ฐœ์˜ ๊ฐ™์€ ์ƒ‰์„ ๊ฐ€์ง„ ํ”ฝ์…€ ์ •๋ณด๋ฅผ ๋‹ค ๋ณด๋‚ด๋Š” ๋Œ€์‹ , ์ƒ‰ ์ •๋ณด์™€ ๋ฐ˜๋ณต๋˜๋Š” ๊ฐœ์ˆ˜๋งŒ ์ „์†กํ•œ๋‹ค.
  • temporal coding example : i+1 ๋ฒˆ์งธ ์ด๋ฏธ์ง€ ํ”ฝ์…€์„ ๋‹ค ๋ณด๋‚ด๋Š”๊ฒŒ ์•„๋‹ˆ๋ผ frame i ์™€ ๋‹ฌ๋ผ์ง„ ๋ถ€๋ถ„๋งŒ ๋ณด๋‚ธ๋‹ค.

Streaming Multimedia : DASH

DASH : Dynamic, Adaptive Streaming over Http

  • Server
    1. ๋น„๋””์˜ค ํŒŒ์ผ์„ ์—ฌ๋Ÿฌ๊ฐœ์˜ ๋ฉ์–ด๋ฆฌ๋กœ ๋‚˜๋ˆˆ๋‹ค
    2. ๊ฐ๊ฐ์˜ ๋ฉ์–ด๋ฆฌ๋ฅผ ์ €์žฅํ•˜๊ณ  ๊ฐ๊ธฐ ๋‹ค๋ฅธ ์†๋„๋กœ ์ธ์ฝ”๋”ฉํ•œ๋‹ค.
    3. Manifest file: ๊ฐ๊ฐ ๋‹ค๋ฅธ ๋ฉ์–ด๋ฆฌ์— ๋Œ€ํ•œ URLs ์„ ์ œ๊ณตํ•œ๋‹ค.์ปดํ“จํŒ…์—์„œ ์ง‘ํ•ฉ์˜ ์ผ๋ถ€ ๋˜๋Š” ๋…ผ๋ฆฌ์ •์—ฐํ•œ ๋‹จ์œ„์ธ ํŒŒ์ผ๋“ค์˜ ๊ทธ๋ฃน์„ ์œ„ํ•œ ๋ฉ”ํƒ€๋ฐ์ดํ„ฐ๋ฅผ ํฌํ•จํ•˜๋Š” ํŒŒ์ผ์ด๋‹ค.
  • Client
    1. ์ฃผ๊ธฐ์ ์œผ๋กœ server-to-client์˜ ๋Œ€์—ญํญ์„ ์ธก์ •ํ•œ๋‹ค.
    2. manifest๋ฅผ consultingํ•˜๊ณ  ํ•œ ๋ฒˆ์— ํ•˜๋‚˜์˜ chunk๋ฅผ ์š”์ฒญํ•œ๋‹ค.
      • ํ˜„์žฌ ์ฃผ์–ด์ง„ ๋Œ€์—ญํญ ์•ˆ์—์„œ ๋ฐ›์„ ์ˆ˜ ์žˆ๋Š” ๊ฐ€์žฅ ํฐ coding rate์˜ ํŒŒ์ผ์„ ๊ณ ๋ฅธ๋‹ค.
      • ์‹œ์ ์— ๋”ฐ๋ผ(์‚ฌ์šฉ ๊ฐ€๋Šฅํ•œ ๋Œ€์—ญํญ์— ๋”ฐ๋ผ) ๋‹ค๋ฅธ coding rate๋ฅผ ์„ ํƒํ•  ์ˆ˜ ์žˆ๋‹ค.
  • Client ์˜ "Intelligence" ; ํด๋ผ์ด์–ธํŠธ๊ฐ€ ๊ฒฐ์ •ํ•˜๋Š” ๊ฒƒ
    1. ์–ธ์ œ chunk๋ฅผ ์š”์ฒญํ• ์ง€ (๋ฒ„ํผ๊ฐ€ ๋‚จ๊ฑฐ๋‚˜ ์˜ค๋ฒ„ํ”Œ๋กœ์šฐ๊ฐ€ ์ผ์–ด๋‚˜์ง€ ์•Š์„ ๋•Œ)
    2. ์–ด๋–ค ์ธ์ฝ”๋”ฉ rate๋ฅผ ์š”์ฒญํ•ด์•ผ ํ• ์ง€ (๋” ๋„“์€ ๋Œ€์—ญํญ์„ ์‚ฌ์šฉ ๊ฐ€๋Šฅํ•  ๊ฒฝ์šฐ ๋” ๋†’์€ ํ’ˆ์งˆ ์ถ”๊ตฌ)
    3. ์–ด๋””๋กœ chunk๋ฅผ ์š”์ฒญํ•ด์•ผ ํ•˜๋Š”์ง€ ( client์—๊ฒŒ ๋” ๊ฐ€๊ฐ‘๊ฑฐ๋‚˜ ๋” ๋งŽ์€ ๋Œ€์—ญํญ์„ ์‚ฌ์šฉ ๊ฐ€๋Šฅํ•œ URL server์— ์š”์ฒญ์„ ๋ณด๋‚ผ ์ˆ˜ ์žˆ๋‹ค)

Case Study : Netflix

2. Content distribution networks ( = CDN )

  • Challenge : ์–ด๋–ป๊ฒŒ ํ•˜๋ฉด ์ˆ˜์‹ญ๋งŒ ๋ช…์˜ ๋™์‹œ ์‚ฌ์šฉ์ž์—๊ฒŒ ์ปจํ…์ธ ๋ฅผ ์ŠคํŠธ๋ฆฌ๋ฐ ํ•  ์ˆ˜ ์žˆ์„๊นŒ?

OPTION 1 : ํ•˜๋‚˜์˜ ๋„“์€ "Mega-server"๋ฅผ ์ด์šฉํ•œ๋‹ค.

  • ํ•˜๋‚˜ ์ž˜๋ชป๋˜๋ฉด ๋‹ค ์ž˜๋ชป๋จ (Single point of failure)
  • network congestion ; ๋„คํŠธ์›Œํฌ์˜ ๋ฐ€์ง‘, ์ •์ฒด
  • ๋จผ ๊ณ ๊ฐ๊นŒ์ง€ ๊ฐ€๋Š” ๊ธธ์ด ๋ฉ€๋‹ค
  • ์—ฌ๋Ÿฌ ๊ฐœ์˜ ๋น„๋””์˜ค์˜ ๋ณต์‚ฌ๋ณธ์ด ์†ก์‹  ๋งํฌ๋ฅผ ํ†ตํ•ด ์ „์†ก๋จ
  • ... ์ด๊ฑด ์ข€ ์•„๋‹Œ ๋“ฏ

OPTION 2 : ์—ฌ๋Ÿฌ ๊ฐœ์˜ ๋น„๋””์˜ค์˜ ๋ณต์‚ฌ๋ณธ์„ ์ง€๋ฆฌํ•™์ ์œผ๋กœ ๋ถ„์‚ฐ๋œ ์—ฌ๋Ÿฌ ๊ณณ์— ๋‚˜๋ˆ„์–ด ์ €์žฅํ•˜๊ณ  ๊ด€๋ฆฌํ•œ๋‹ค.

  • CDN ์„œ๋ฒ„๋ฅผ ์—ฌ๋Ÿฌ access networks(์‚ฌ์šฉ์ž๋“ค์—๊ฒŒ ๊ฐ€๊นŒ์šด)์— ๋ฐ•์•„ ๋„ฃ์–ด๋‘”๋‹ค.
  • Smaller number (10’s) of larger clusters in POPs (points of presence) near (but not within) access networks
  • - limelight์— ์˜ํ•ด ์‚ฌ์šฉ๋จ
  • More Challenges
    1. ์–ด๋–ค CDN node์—์„œ ์ปจํ…์ธ ๋ฅผ ๊ฒ€์ƒ‰ํ•ด์˜ฌ ๊ฒƒ์ธ๊ฐ€
    2. ์–ด๋–ค ์ปจํ…์ธ ๋ฅผ ์–ด๋–ค CDN node์— ๋ฐฐ์น˜ํ•  ๊ฒƒ์ธ๊ฐ€

CDN Content Access : A closer Look

Socket programming

1. UDP socket ; unreliable datagram

  • ํด๋ผ์ด์–ธํŠธ์™€ ์„œ๋ฒ„ ์‚ฌ์ด์— No Connection
    1. ๋ฐ์ดํ„ฐ ์ „์†ก ์ „์— handshaking ๋‹จ๊ณ„ X
    2. Sender๋Š” ๋„์ฐฉ์ง€์˜ IP ์ฃผ์†Œ์™€ ํฌํŠธ ๋„˜๋ฒ„๋ฅผ ํŒจํ‚ท์— ๋ช…์‹œ์ ์œผ๋กœ ๋‹ด์•„์„œ ๋ณด๋‚ธ๋‹ค
    3. Receiver๋Š” ์ˆ˜์‹ ํ•œ ํŒจํ‚ท์—์„œ sender์˜ IP ์ฃผ์†Œ์™€ ํฌํŠธ ๋„˜๋ฒ„๋ฅผ ์ถ”์ถœํ•œ๋‹ค.
    4. ์ „์†ก๋œ ๋ฐ์ดํ„ฐ๊ฐ€ ๋ถ„์‹ค๋˜๊ฑฐ๋‚˜, ์š”์ฒญ์ด ์ง€๋‚œ ํ›„ ๋ฐ›๊ฒŒ๋  ์ˆ˜ ์žˆ๋‹ค.
  • *UDP๋Š” ํด๋ผ์ด์–ธํŠธ์™€ ์„œ๋ฒ„ ๊ฐ„์— ์‹ ๋ขฐํ•  ์ˆ˜ ์—†๋Š” bytes ๊ทธ๋ฃน("datagram") ์ „์†ก ๊ธฐ๋Šฅ์„ ์ œ๊ณตํ•œ๋‹ค. *

2. TCP socket ; reliable, byte stream-oriented

*TCP๋Š” ํด๋ผ์ด์–ธํŠธ์™€ ์„œ๋ฒ„ ๊ฐ„์— ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š”, ์ˆœ์„œ๊ฐ€ ์žˆ๋Š” byte-stream ์ „์†ก("PIPE") ๋ฅผ ์ œ๊ณตํ•œ๋‹ค. *

  • ํด๋ผ์ด์–ธํŠธ๋Š” ๋ฌด์กฐ๊ฑด ์„œ๋ฒ„์™€ contact ํ•ด์•ผ ํ•จ!!
    1. ์„œ๋ฒ„ ํ”„๋กœ์„ธ์Šค๊ฐ€ ์ผ๋‹จ ๋จผ์ € ์‹คํ–‰ ๋˜์–ด์•ผ ๋ญ˜ ํ•จ
    2. ์„œ๋ฒ„๋Š” client์˜ contact์„ ๋ฐ›์„ socket์„ ๋งŒ๋“ค๊ฒŒ ๋จ
  • ํด๋ผ์ด์–ธํŠธ๊ฐ€ ์„œ๋ฒ„์—๊ฒŒ contact ํ•˜๋ ค๋ฉด
    1. TCP Socket์„ ๋งŒ๋“ฆ (์„œ๋ฒ„ ํ”„๋กœ์„ธ์Šค์˜ IP address์™€ port number๋กœ ๊ตฌ๋ถ„๋˜๋Š”)
    2. client๊ฐ€ socket์„ ๋งŒ๋“ค ๋•Œ client TCP๋Š” server TCP์—๊ฒŒ ์—ฐ๊ฒฐ์„ ์„ค์ •ํ•จ(establishes)
  • ์„œ๋ฒ„๋Š” client์—๊ฒŒ contacted ๋˜๋ฉด ์ƒˆ๋กœ์šด TCP socket์„ ๋งŒ๋“ฆ
    1. ๊ทธ client์™€ ํ†ต์‹ ํ•˜๊ธฐ ์œ„ํ•ด!
    2. ์„œ๋ฒ„๊ฐ€ ์—ฌ๋Ÿฌ ํด๋ผ์ด์–ธํŠธ์™€ ๋Œ€ํ™”ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•˜๋ ค๊ณ !
    3. Source port number๋Š” ๊ฐ ํด๋ผ์ด์–ธํŠธ๋“ค์„ ๊ตฌ๋ถ„ํ•˜๋Š”๋ฐ ์“ฐ์ธ๋‹ค.