Skip to content

Attacks on Networks and Communication Paths

頻出~48 min7 terms9 past questions

Goals of this unit

  • Sort attacks that cut into communication by where they cut in: partway along the network path, or inside the user's browser.
  • For abuse of mail relaying (third-party relay, also called open relay), explain the act of abuse and the server configuration that allows it as two sides of the same thing.
  • Distinguish spoofing and hijacking techniques by the information they fake or steal: the source IP address, name-resolution information, or a session ID.
Terms at a glance (7)
TermIn one lineWhat is faked or abusedWhere it happens
中間者攻撃Gets between two communicating parties to read or rewrite what they exchangeData flowing between the two partiesPartway along the network path
MITB攻撃Malware hiding in the browser rewrites the screen and outgoing data from insideTransaction details shown and entered in the browserInside the user's browser
第三者中継An unrelated party gets someone else's mail server to send their mailA mail server that does not restrict relayingThe mail server
オープンリレーA server configuration that relays mail from anyone to anyone without restrictionThe configuration itself, which places no limit on where mail is relayedThe mail server
IPスプーフィングForges a packet's source IP address to pose as another deviceA packet's source IP addressPackets on the network
DNSキャッシュポイズニングMakes a DNS server remember a false mapping and sends users to a fake siteThe mapping between domain names and IP addressesThe DNS server's cache
セッションハイジャックSteals the information that identifies post-login traffic and carries on as the userThe session IDPost-login communication

Terms

中間者攻撃Man-in-the-Middle (MITM) Attack

Open in glossary 中間者攻撃

A man-in-the-middle attack (中間者攻撃) is one in which an attacker secretly places themselves between two communicating parties and eavesdrops (盗聴) on or tampers (改ざん) with the data they exchange.

利用者

攻撃者(中間者)

サーバ

The attacker poses as the server to the user and as the user to the server, passing data in both directions through their own hands. Because each side believes it is talking directly to the other, neither is likely to notice when the contents are read or altered on the way. Unencrypted traffic and connections made without checking whether the other side is genuine are easy targets, and the defenses are encrypting the traffic and confirming the other party is authentic, for instance by validating its server certificate.

AnalogyIt is like a messenger carrying letters between two people who quietly opens each one, reads it, adds a few lines, and then delivers it. The recipient has no idea the messenger tampered with it along the way.

How to rememberThe Japanese name is a direct rendering of "man-in-the-middle", shortened to MITM from its initials.

Exam points

  • Mix-upA man-in-the-middle attack (中間者攻撃) goes beyond eavesdropping (盗聴), which only reads the data: the attacker can also rewrite it while passing it on.

1 past questions on this term 中間者攻撃

MITB攻撃Man-in-the-Browser (MITB) Attack

Open in glossary MITB攻撃

A man-in-the-browser (MITB) attack (MITB攻撃) is one in which malware (マルウェア, malicious software) that has got into the user's device takes over the browser and rewrites what the screen shows and what gets sent, from inside the browser.

  1. Malware hides in the browser
  2. The user enters the transaction
  3. The content is altered before sending
  4. It reaches the bank over the genuine connection

The altered request arrives over the genuine encrypted connection, and the bank processes it as the customer's own.

4 / 4

The main target is online banking: the site the user opens is the real one, and the user logs in correctly themselves. The malware may even keep showing the details exactly as the user typed them, so the user is unlikely to notice anything wrong.

AnalogyIt is like a fake clerk standing beside you at the counter who swaps your completed application form for a different one just as it leaves your hand. The counter accepts the form it receives as yours.

How to rememberThe name swaps the "middle" in man-in-the-middle for "browser": the man in the middle has moved into the browser.

Exam points

  • ExamIn a man-in-the-browser attack (MITB攻撃), malware (マルウェア) that has hijacked the browser quietly alters online banking (インターネットバンキング) transactions without the user noticing, leading to fraudulent transfers and the like.
  • Mix-upA man-in-the-browser attack (MITB攻撃) alters data on the device before it is encrypted, so encrypting (暗号化) the connection does not stop it. What helps is protection against malware (マルウェア) on the device and a way to confirm the transaction through a separate channel.

1 past questions on this term MITB攻撃

Man-in-the-Middle (MITM) Attack vs. Man-in-the-Browser (MITB) Attack

AspectMan-in-the-Middle (MITM) AttackMan-in-the-Browser (MITB) Attack
Key differenceWhere the attacker gets inPartway along the network pathInside the user's browser
Who carries it outAn attacker who has cut into the pathMalware that has got into the device
Encrypting the connectionEffective when combined with verifying the other partyUseless, since data is altered before encryption

第三者中継Third-Party Relay

Open in glossary 第三者中継

Third-party relay (第三者中継) is the act of an outsider with no connection to a mail server getting that server to relay (中継, forward a received message on to another server) their mail, using it to send messages.

  1. An outside attacker asks it to send mail
  2. An unrestricted mail server accepts
  3. Relayed to outside recipients
  4. The server is seen as the sender

To the recipients, the relaying server looks like the sender, and the attacker stays hidden.

4 / 4

Attackers typically look for servers that accept mail from outside addressed to outside, route mail through them, and send spam in bulk. A device used as a pass-through for abuse like this is called a stepping stone (踏み台); because the recipients see the stepping-stone server as the sender, the attacker can hide their own identity.

AnalogyIt is like a company reception desk taking a stack of letters from a passer-by who is neither staff nor a visitor and mailing them out under the company's name. The complaints go to the company, not the passer-by.

Exam points

  • Mix-upThird-party relay (第三者中継) does not only harm the people who receive the spam (迷惑メール). The owner of the mail server it passes through is also taken for the sender and loses trust.

オープンリレーOpen Relay

Open in glossary オープンリレー

An open relay (オープンリレー) is a mail server configuration that relays mail from anyone to anyone without restriction, never checking the sender or the recipient; it is also called third-party relay (第三者中継).

Mail flowProper configurationOpen relay
From the organization's users to outsideRelayedRelayed
From outside to the organization's usersAcceptedAccepted
From outside to outsideRejectedRelayed anyway

Once attackers find an open relay, they use it as a stepping stone for third-party relay. The fix is to configure the server to relay only mail sent by the organization's own users and to reject mail from outside addressed to outside. This setting still lets the organization's users send mail and still accepts mail from outside addressed to the organization.

AnalogyIt is like a parcel depot that sends out anything brought to it without looking at either the address or the sender.

How to remember"Open" means open to anyone, and "relay" is the same word as in a relay race, where the baton is handed on to the next runner.

Exam points

  • Mix-upOpen relay (オープンリレー) and third-party relay (第三者中継) are often used to mean the same thing. The first names it from the side of the server's configuration weakness, the second from the side of whoever exploits that weakness to send mail through someone else's server.

Third-Party Relay vs. Open Relay

AspectThird-Party RelayOpen Relay
Key differenceMain focusThe act of sending mail through someone else's serverA server configuration that relays without restriction
How they relateThe result of exploiting the weaknessThe cause that permits the abuse
Whose side it describesThe outsider abusing the serverThe server being abused

IPスプーフィングIP Spoofing

Open in glossary IPスプーフィング

IP spoofing (IPスプーフィング) is a technique that forges the source IP address (IPアドレス) of the packets (パケット, the small units data is split into for sending) so they look as if they came from another device.

  1. The source field is rewritten to a permitted device's address
  2. A device that judges senders by IP address alone
  3. Lets it through as permitted traffic

Each packet carries a source IP address alongside its destination, and the receiving device normally trusts that value to decide who sent it. Because the sender can rewrite the source value, a system that permits traffic on IP address alone can be breached by packets posing as an approved device, and the countermeasure is authentication that does not depend only on the IP address. The technique is also used to hide where an attack comes from.

AnalogyIt is like writing a trusted business partner's address in the return-address corner of an envelope. If the recipient opens it based on the return address alone, they are fooled.

How to rememberThe English word "spoof" means to trick someone or to mimic something in fun.

Exam points

  • ExamIP spoofing (IPスプーフィング) is a means of unauthorized access that sends packets (パケット) with a forged source IP address (IPアドレス).
  • ExamIP spoofing (IPスプーフィング) is a technical disguise aimed at machines. It is not social engineering (ソーシャルエンジニアリング), which extracts information by playing on people's psychology and carelessness.

1 past questions on this term IPスプーフィング

DNSキャッシュポイズニングDNS Cache Poisoning

Open in glossary DNSキャッシュポイズニング

DNS cache poisoning (DNSキャッシュポイズニング) is an attack that slips false entries into the name-resolution information a DNS server (DNSサーバ) keeps temporarily, so that users are sent to a fake site.

  1. The attacker injects a false mapping
  2. It stays in the DNS server's cache
  3. The user queries the correct domain name
  4. A false IP address comes back and leads to the fake site

The false IP address in the cache is returned, and the user's browser connects to the fake site.

4 / 4

DNS is the name-resolution system that looks up a device's IP address from its domain name, and a DNS server keeps the results it has looked up as a cache (キャッシュ, temporary storage for reuse). While the cache is poisoned, every user who asks that DNS server about the domain name is affected. This can lead to login details and other information being entered on the fake site and stolen, and DNSSEC, which uses digital signatures to verify that DNS responses are genuine, is one countermeasure.

AnalogyIt is like someone writing a fake shop's address onto the map at a town information desk as the location of a real shop. The people who ask for the shop by its correct name are exactly the ones sent straight to the fake one.

How to remember"Poisoning" pictures slipping poison into the cache, the pool where name-resolution information collects.

Exam points

  • ExamDNS cache poisoning (DNSキャッシュポイズニング) rewrites, on a server, the information used to look up an IP address (IPアドレス) from a domain name (ドメイン名), so that users land on a fake site.
  • Mix-upPhishing (フィッシング) tricks people with fake emails and the like into visiting a fake site. DNS cache poisoning (DNSキャッシュポイズニング) corrupts name-resolution information instead, so users reach the fake site even when they type the correct domain name.

1 past questions on this term DNSキャッシュポイズニング

IP Spoofing vs. DNS Cache Poisoning

AspectIP SpoofingDNS Cache Poisoning
Key differenceWhat is fakedA packet's source IP addressThe mapping between domain names and IP addresses
Who accepts the false informationThe device receiving the packetThe DNS server answering queries
Main aimPosing as another device, hiding the sourceSending users to a fake site

セッションハイジャックSession Hijacking

Open in glossary セッションハイジャック

Session hijacking (セッションハイジャック) is an attack in which the attacker steals the information that identifies the continuing exchange after login, the session (セッション), and takes over that communication by posing as the user.

  1. A session ID is issued at login
  2. The attacker obtains the session ID
  3. The attacker sends requests with that ID
  4. The server treats them as the user's own

The server does not ask for the password again and accepts the attacker's requests as the user's.

4 / 4

A web service issues each logged-in user an identifying value called a session ID (セッションID) and passes it back and forth in a cookie (a small piece of data a site has the browser store) or similar, to confirm the user is logged in. If an attacker obtains this value by eavesdropping, guessing, or extracting it with a script, they can act as the user without knowing the password.

AnalogyIt is like stealing the wristband an amusement park puts on you at the gate: with it, someone can ride the attractions as you without buying a ticket. The staff let people through by looking at the wristband alone.

How to remember"Hijack" is the same word used for seizing a plane or a car; here the thing seized is the session.

Exam points

  • ExamIf a cookie holding session-identifying information leaks through cross-site scripting (クロスサイトスクリプティング) or a similar attack, the attacker may be treated as the user and misuse that user's account.
  • Mix-upUnlike attacks that steal passwords, session hijacking (セッションハイジャック) seizes the state that exists after login is complete. The attacker gets into the user's communication without passing authentication.

1 past questions on this term セッションハイジャック

Key points

  • In attacks that cut into communication, data is read or rewritten while both the user and the other party believe they are talking directly.
  • Configure mail servers not to relay mail unrelated to the organization, so they are not used as a stepping stone (踏み台) for spam.
  • The sender can rewrite a packet's source IP address, so any system that trusts the other party by IP address alone can be broken.
  • Once a DNS server's cache is poisoned, users reach a fake site even when they type the domain name correctly.
  • If the session ID (セッションID) is stolen, an attacker who does not know the password is treated as the logged-in user.

Check questions

5 past questions from this unit

Start the check questions

All past questions in this topic